Particle Swarm Optimization for SineWave Dataset.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Particle Swarm Optimization for SineWave Dataset"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Import Libraries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 357,
   "metadata": {},
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "\n",
    "import pandas as pd \n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.model_selection import train_test_split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 358,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Import Keras\n",
    "from keras import Sequential\n",
    "from keras.layers import LSTM, Dropout, Dense, TimeDistributed, Activation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 359,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Import PySwarms\n",
    "import pyswarms as ps\n",
    "from pyswarms.utils.plotters import (plot_cost_history)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Load dataset\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 360,
   "metadata": {},
   "outputs": [],
   "source": [
    "sineData = pd.read_csv('./Sinewave.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 361,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>sinewave</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>5001.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>0.000168</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.707207</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>-0.999986</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>-0.688408</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>0.005310</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>0.725324</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>0.999986</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          sinewave\n",
       "count  5001.000000\n",
       "mean      0.000168\n",
       "std       0.707207\n",
       "min      -0.999986\n",
       "25%      -0.688408\n",
       "50%       0.005310\n",
       "75%       0.725324\n",
       "max       0.999986"
      ]
     },
     "execution_count": 361,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sineData.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 362,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>sinewave</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.841471</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.873736</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.902554</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.927809</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.949402</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   sinewave\n",
       "0  0.841471\n",
       "1  0.873736\n",
       "2  0.902554\n",
       "3  0.927809\n",
       "4  0.949402"
      ]
     },
     "execution_count": 362,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sineData.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 363,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(5001, 1)"
      ]
     },
     "execution_count": 363,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sineData.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Split data to train and test set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 364,
   "metadata": {},
   "outputs": [],
   "source": [
    "sine_values = []\n",
    "steps = 1\n",
    "input_width = 5\n",
    "\n",
    "x = []\n",
    "y = []\n",
    "\n",
    "for i in range(0, sineData.shape[0], steps):\n",
    "    sine_values.append(sineData.at[i,'sinewave'])\n",
    "\n",
    "for i in range(0, len(sine_values) - input_width, steps): \n",
    "    x.append(sine_values[i: i + input_width])\n",
    "    y.append(sine_values[i + input_width])\n",
    "\n",
    "x = np.reshape(x, [-1, input_width, 1])\n",
    "y = np.reshape(y, [-1, 1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 365,
   "metadata": {},
   "outputs": [],
   "source": [
    "trainX, testX, trainY, testY = train_test_split(x, y, test_size=0.20, shuffle=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 366,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(3996, 5, 1)"
      ]
     },
     "execution_count": 366,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "trainX.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 367,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1000, 5, 1)"
      ]
     },
     "execution_count": 367,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "testX.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Create neural network model using backpropagation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 368,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = Sequential()\n",
    "model.add(LSTM(32, return_sequences=False, input_shape=(None, 1)))\n",
    "model.add(Dense(1, activation='linear'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 442,
   "metadata": {},
   "outputs": [],
   "source": [
    "# https://github.com/keras-team/keras/issues/7947\n",
    "def rmse_nn(y_true, y_pred):\n",
    "    from keras import backend\n",
    "    return backend.sqrt(backend.mean(backend.square(y_pred - y_true), axis=-1))\n",
    "\n",
    "def r_square_loss(y_true, y_pred):\n",
    "    from keras import backend as K\n",
    "    SS_res =  K.sum(K.square(y_true - y_pred)) \n",
    "    SS_tot = K.sum(K.square(y_true - K.mean(y_true))) \n",
    "    return 1 - ( 1 - SS_res/(SS_tot + K.epsilon()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 443,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "lstm_9 (LSTM)                (None, 32)                4352      \n",
      "_________________________________________________________________\n",
      "dense_8 (Dense)              (None, 1)                 33        \n",
      "=================================================================\n",
      "Total params: 4,385\n",
      "Trainable params: 4,385\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model.compile(loss=rmse_nn, optimizer='sgd', metrics=[rmse_nn, r_square_loss])\n",
    "model.summary()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Train Model using backpropagation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 444,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 3596 samples, validate on 400 samples\n",
      "Epoch 1/20\n",
      "3596/3596 [==============================] - 8s 2ms/step - loss: 0.0852 - rmse_nn: 0.0852 - r_square_loss: 0.0191 - val_loss: 0.0851 - val_rmse_nn: 0.0851 - val_r_square_loss: 0.0233\n",
      "Epoch 2/20\n",
      "3596/3596 [==============================] - 2s 431us/step - loss: 0.0846 - rmse_nn: 0.0846 - r_square_loss: 0.0190 - val_loss: 0.0841 - val_rmse_nn: 0.0841 - val_r_square_loss: 0.0226\n",
      "Epoch 3/20\n",
      "3596/3596 [==============================] - 1s 411us/step - loss: 0.0838 - rmse_nn: 0.0838 - r_square_loss: 0.0184 - val_loss: 0.0832 - val_rmse_nn: 0.0832 - val_r_square_loss: 0.0223\n",
      "Epoch 4/20\n",
      "3596/3596 [==============================] - 1s 397us/step - loss: 0.0830 - rmse_nn: 0.0830 - r_square_loss: 0.0181 - val_loss: 0.0826 - val_rmse_nn: 0.0826 - val_r_square_loss: 0.0218\n",
      "Epoch 5/20\n",
      "3596/3596 [==============================] - 2s 479us/step - loss: 0.0822 - rmse_nn: 0.0822 - r_square_loss: 0.0178 - val_loss: 0.0816 - val_rmse_nn: 0.0816 - val_r_square_loss: 0.0212\n",
      "Epoch 6/20\n",
      "3596/3596 [==============================] - 1s 403us/step - loss: 0.0812 - rmse_nn: 0.0812 - r_square_loss: 0.0174 - val_loss: 0.0809 - val_rmse_nn: 0.0809 - val_r_square_loss: 0.0211\n",
      "Epoch 7/20\n",
      "3596/3596 [==============================] - 1s 379us/step - loss: 0.0806 - rmse_nn: 0.0806 - r_square_loss: 0.0168 - val_loss: 0.0798 - val_rmse_nn: 0.0798 - val_r_square_loss: 0.0202\n",
      "Epoch 8/20\n",
      "3596/3596 [==============================] - 1s 413us/step - loss: 0.0796 - rmse_nn: 0.0796 - r_square_loss: 0.0167 - val_loss: 0.0789 - val_rmse_nn: 0.0789 - val_r_square_loss: 0.0198\n",
      "Epoch 9/20\n",
      "3596/3596 [==============================] - 2s 427us/step - loss: 0.0786 - rmse_nn: 0.0786 - r_square_loss: 0.0162 - val_loss: 0.0782 - val_rmse_nn: 0.0782 - val_r_square_loss: 0.0198\n",
      "Epoch 10/20\n",
      "3596/3596 [==============================] - 1s 417us/step - loss: 0.0778 - rmse_nn: 0.0778 - r_square_loss: 0.0159 - val_loss: 0.0772 - val_rmse_nn: 0.0772 - val_r_square_loss: 0.0191\n",
      "Epoch 11/20\n",
      "3596/3596 [==============================] - 1s 369us/step - loss: 0.0769 - rmse_nn: 0.0769 - r_square_loss: 0.0153 - val_loss: 0.0764 - val_rmse_nn: 0.0764 - val_r_square_loss: 0.0187\n",
      "Epoch 12/20\n",
      "3596/3596 [==============================] - 2s 531us/step - loss: 0.0761 - rmse_nn: 0.0761 - r_square_loss: 0.0153 - val_loss: 0.0753 - val_rmse_nn: 0.0753 - val_r_square_loss: 0.0181\n",
      "Epoch 13/20\n",
      "3596/3596 [==============================] - 2s 585us/step - loss: 0.0751 - rmse_nn: 0.0751 - r_square_loss: 0.0148 - val_loss: 0.0743 - val_rmse_nn: 0.0743 - val_r_square_loss: 0.0176\n",
      "Epoch 14/20\n",
      "3596/3596 [==============================] - 2s 460us/step - loss: 0.0741 - rmse_nn: 0.0741 - r_square_loss: 0.0143 - val_loss: 0.0734 - val_rmse_nn: 0.0734 - val_r_square_loss: 0.0172\n",
      "Epoch 15/20\n",
      "3596/3596 [==============================] - 1s 374us/step - loss: 0.0731 - rmse_nn: 0.0731 - r_square_loss: 0.0140 - val_loss: 0.0724 - val_rmse_nn: 0.0724 - val_r_square_loss: 0.0168\n",
      "Epoch 16/20\n",
      "3596/3596 [==============================] - 1s 411us/step - loss: 0.0721 - rmse_nn: 0.0721 - r_square_loss: 0.0137 - val_loss: 0.0716 - val_rmse_nn: 0.0716 - val_r_square_loss: 0.0161\n",
      "Epoch 17/20\n",
      "3596/3596 [==============================] - 1s 401us/step - loss: 0.0710 - rmse_nn: 0.0710 - r_square_loss: 0.0132 - val_loss: 0.0703 - val_rmse_nn: 0.0703 - val_r_square_loss: 0.0156\n",
      "Epoch 18/20\n",
      "3596/3596 [==============================] - 2s 535us/step - loss: 0.0698 - rmse_nn: 0.0698 - r_square_loss: 0.0129 - val_loss: 0.0700 - val_rmse_nn: 0.0700 - val_r_square_loss: 0.0154\n",
      "Epoch 19/20\n",
      "3596/3596 [==============================] - 2s 451us/step - loss: 0.0689 - rmse_nn: 0.0689 - r_square_loss: 0.0125 - val_loss: 0.0686 - val_rmse_nn: 0.0686 - val_r_square_loss: 0.0147\n",
      "Epoch 20/20\n",
      "3596/3596 [==============================] - 2s 470us/step - loss: 0.0679 - rmse_nn: 0.0679 - r_square_loss: 0.0119 - val_loss: 0.0671 - val_rmse_nn: 0.0671 - val_r_square_loss: 0.0142\n"
     ]
    }
   ],
   "source": [
    "historyNN = model.fit(trainX, trainY, batch_size=50, epochs=20, validation_split=0.1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Model Training Performance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 454,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(historyNN.history['loss'])\n",
    "plt.plot(historyNN.history['val_loss'])\n",
    "plt.title('Model Loss Plot')\n",
    "plt.ylabel('Loss')\n",
    "plt.xlabel('Epoch')\n",
    "plt.legend(['Train', 'Validation'], loc='upper left')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Test Neural Network Model - BackPropagation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 452,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 0s 146us/step\n"
     ]
    }
   ],
   "source": [
    "lossNN = model.evaluate(testX, testY)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 453,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The loss output for test dataset is: 0.067104\n"
     ]
    }
   ],
   "source": [
    "print('The loss output for test dataset is: %f' % lossNN[0])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Plot the sine wave from the test set prediction vs actual"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 375,
   "metadata": {},
   "outputs": [],
   "source": [
    "predictY = model.predict(testX)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 376,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "time = np.arange(predictY.shape[0])\n",
    "\n",
    "plt.plot(time, predictY, testY)\n",
    "plt.xlabel('time')\n",
    "plt.ylabel('amplitude')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Part 2 - Neural Network using GA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 377,
   "metadata": {},
   "outputs": [],
   "source": [
    "modelGA = Sequential()\n",
    "modelGA.add(LSTM(32, return_sequences=False, input_shape=(None, 1), name='layer_1'))\n",
    "modelGA.add(Dense(1, activation='linear', name='layer_2'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 378,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "layer_1 (LSTM)               (None, 32)                4352      \n",
      "_________________________________________________________________\n",
      "layer_2 (Dense)              (None, 1)                 33        \n",
      "=================================================================\n",
      "Total params: 4,385\n",
      "Trainable params: 4,385\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "modelGA.compile(loss=\"mean_squared_error\", optimizer='sgd', metrics=[\"mean_squared_error\", r_square_loss])\n",
    "modelGA.summary()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Loss function for Pyswarm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 445,
   "metadata": {},
   "outputs": [],
   "source": [
    "# root mean squared error (rmse) for regression\n",
    "def rmse(y_true, y_pred):\n",
    "    return np.sqrt(np.mean(np.square(y_pred - y_true), axis=-1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 428,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Get the loss value for each particle\n",
    "def getFitness(params):\n",
    "    layer_1 = modelGA.get_layer('layer_1').get_weights()\n",
    "    layer_2 = modelGA.get_layer('layer_2').get_weights()\n",
    "    \n",
    "    layer1_W1_shape = (layer_1[0].shape[0] * layer_1[0].shape[1])\n",
    "    layer1_W2_shape = layer1_W1_shape + (layer_1[1].shape[0] * layer_1[1].shape[1])\n",
    "    layer1_B_shape = layer1_W2_shape + (layer_1[2].shape[0])\n",
    "    layer2_W_shape = layer1_B_shape + (layer_2[0].shape[0] * layer_2[0].shape[1])\n",
    "    layer2_B_shape = layer2_W_shape + (layer_2[1].shape[0])\n",
    "\n",
    "    layer1_W1 = params[0:layer1_W1_shape].reshape(layer_1[0].shape)\n",
    "    layer1_W2 = params[layer1_W1_shape:layer1_W2_shape].reshape(layer_1[1].shape)\n",
    "    layer1_B = params[layer1_W2_shape:layer1_B_shape].reshape(layer_1[2].shape)\n",
    "    layer2_W = params[layer1_B_shape:layer2_W_shape].reshape(layer_2[0].shape)\n",
    "    layer2_B = params[layer2_W_shape:layer2_B_shape].reshape(layer_2[1].shape)\n",
    "    \n",
    "    modelGA.get_layer('layer_1').set_weights([layer1_W1, layer1_W2, layer1_B])\n",
    "    modelGA.get_layer('layer_2').set_weights([layer2_W, layer2_B])\n",
    "    \n",
    "    predY = modelGA.predict(trainX)\n",
    "    loss = rmse(trainY.reshape(trainY.shape[0]), predY.reshape(trainY.shape[0]))\n",
    "    return loss"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 429,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Computed loss for all particles\n",
    "def f(params):\n",
    "    print('Number of particles: %d' % params.shape[0])\n",
    "    losses = np.array([getFitness(params[i]) for i in range(params.shape[0])])\n",
    "    print('List of losses for all particles')\n",
    "    print(losses)\n",
    "    return losses"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Pyswarm algorithm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 430,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2019-03-02 14:58:09,315 - pyswarms.single.global_best - INFO - Optimize for 50 iters with {'c1': 0.5, 'c2': 0.3, 'w': 0.9}\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:   0%|          |0/50\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of params in Neural Network: 4385\n",
      "Number of particles: 20\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:   0%|          |0/50, best_cost=13\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:   2%|▏         |1/50, best_cost=13\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "List of losses for all particles\n",
      "[16.09386572 14.04286392 15.72491667 15.26946004 17.6489134  16.23777086\n",
      " 14.99025789 14.24477598 14.81692274 15.86529524 15.8349588  14.64303822\n",
      " 14.8896217  15.62132171 14.66467107 16.77240304 15.28026794 13.04812632\n",
      " 15.75681363 16.35977631]\n",
      "Number of particles: 20\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:   2%|▏         |1/50, best_cost=13\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:   4%|▍         |2/50, best_cost=13\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "List of losses for all particles\n",
      "[28.62166195 29.66537364 28.61878187 30.21746838 32.80991469 28.58901165\n",
      " 27.76975213 30.99974571 29.83592465 31.53821402 31.66014496 28.20342063\n",
      " 30.27984993 30.95185591 27.47841644 30.98993045 31.63844929 27.62388906\n",
      " 27.93019965 29.04805017]\n",
      "Number of particles: 20\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:   4%|▍         |2/50, best_cost=13\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:   6%|▌         |3/50, best_cost=13\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "List of losses for all particles\n",
      "[33.53309575 37.19122494 33.02261558 36.09260618 38.9693891  34.27985874\n",
      " 31.80618645 39.28822671 38.12206063 40.878422   38.76594414 34.31722257\n",
      " 37.23340978 37.42092786 33.44302741 36.51196104 38.03862089 34.96276423\n",
      " 33.4398559  34.3124064 ]\n",
      "Number of particles: 20\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:   6%|▌         |3/50, best_cost=13\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:   8%|▊         |4/50, best_cost=13\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "List of losses for all particles\n",
      "[30.73450587 34.68777107 31.14714974 33.75542122 34.77399977 31.31613086\n",
      " 29.34354267 36.03441758 36.79768184 38.47782259 36.65362229 32.85170586\n",
      " 34.65153121 32.96457472 32.07782618 33.09087624 35.40960928 33.02386096\n",
      " 31.7331283  31.86132191]\n",
      "Number of particles: 20\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:   8%|▊         |4/50, best_cost=13\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  10%|█         |5/50, best_cost=13\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "List of losses for all particles\n",
      "[22.43925996 24.97001112 23.66944102 25.68148253 21.46634845 22.7451626\n",
      " 20.91706718 24.08434259 25.42096263 26.49965406 26.81676101 25.15724221\n",
      " 22.44252681 22.7385746  24.94872127 23.57832587 26.47730751 23.26909842\n",
      " 22.83922055 21.65711626]\n",
      "Number of particles: 20\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  10%|█         |5/50, best_cost=7.72\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  12%|█▏        |6/50, best_cost=7.72\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "List of losses for all particles\n",
      "[11.60070142 11.47324575 11.74271596 14.12271549  7.71742983 11.82592793\n",
      "  8.54808304 10.02319568 10.89452924 13.01180659 12.99344687 12.91913605\n",
      "  8.65937463  8.9343844  12.95050492 10.7161654  11.69596856  9.78305061\n",
      " 11.61194575  9.03405748]\n",
      "Number of particles: 20\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  12%|█▏        |6/50, best_cost=0.321\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  14%|█▍        |7/50, best_cost=0.321\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "List of losses for all particles\n",
      "[1.36743728 1.30295098 0.83623998 0.6699191  0.87895415 0.6617977\n",
      " 0.72689144 0.47872521 0.64133319 0.84212772 1.39183853 1.17646437\n",
      " 1.16876805 1.03401326 0.60961263 0.32125167 0.93863757 1.27676485\n",
      " 0.52728269 1.07709422]\n",
      "Number of particles: 20\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  14%|█▍        |7/50, best_cost=0.258\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  16%|█▌        |8/50, best_cost=0.258\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "List of losses for all particles\n",
      "[1.21876105 1.25765175 0.30243248 0.88796643 0.71266407 0.759373\n",
      " 0.84036873 0.36518521 0.65478166 0.92445981 0.90389385 1.11318093\n",
      " 0.81917819 0.95074767 0.45081131 0.52695113 1.3074086  1.63843787\n",
      " 0.25818076 1.45199945]\n",
      "Number of particles: 20\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  16%|█▌        |8/50, best_cost=0.258\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  18%|█▊        |9/50, best_cost=0.258\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "List of losses for all particles\n",
      "[1.10369406 1.40049424 1.54949832 0.76850898 4.39407721 0.56214703\n",
      " 0.92312686 1.29174296 0.708158   0.90620947 1.27845403 1.01393319\n",
      " 0.78041322 2.11517745 1.31118854 0.47233183 1.23996671 1.55071474\n",
      " 1.55328039 1.56091631]\n",
      "Number of particles: 20\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  18%|█▊        |9/50, best_cost=0.258\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  20%|██        |10/50, best_cost=0.258\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "List of losses for all particles\n",
      "[2.63961828 2.73704289 3.33413194 1.51667612 5.70367385 1.72189913\n",
      " 0.6530953  2.67452439 0.74394977 0.84987829 2.81789906 1.71940782\n",
      " 5.31526021 3.23995556 3.36343301 0.63022722 1.57214694 1.51507168\n",
      " 2.2147908  1.29200591]\n",
      "Number of particles: 20\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  20%|██        |10/50, best_cost=0.258\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  22%|██▏       |11/50, best_cost=0.258\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "List of losses for all particles\n",
      "[3.77534357 2.39010989 2.53197418 1.00311632 4.04210604 2.15303317\n",
      " 0.55126989 2.62602271 1.36298676 1.17822362 2.64156971 2.80812424\n",
      " 6.8310495  2.88840634 3.10965329 1.21001348 1.0734312  0.27543506\n",
      " 2.06474886 0.92283813]\n",
      "Number of particles: 20\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  22%|██▏       |11/50, best_cost=0.258\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  24%|██▍       |12/50, best_cost=0.258\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "List of losses for all particles\n",
      "[2.68445812 1.21971994 0.93927518 0.62031735 1.60731121 1.75880731\n",
      " 0.77356871 1.18472942 1.19765033 0.73487284 0.80497405 1.94640578\n",
      " 4.83168182 0.79208078 0.92681349 1.05335008 1.1875363  0.46785745\n",
      " 0.86938041 0.82306723]\n",
      "Number of particles: 20\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  24%|██▍       |12/50, best_cost=0.247\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  26%|██▌       |13/50, best_cost=0.247\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "List of losses for all particles\n",
      "[1.32853529 1.01752189 0.37037145 1.52628638 1.05238589 1.45679137\n",
      " 0.63338641 0.27143627 1.50080253 0.85788031 0.46383346 1.16425185\n",
      " 1.98037068 0.39962089 1.12668382 1.91288731 2.171303   0.29165408\n",
      " 0.24684312 0.28868748]\n",
      "Number of particles: 20\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  26%|██▌       |13/50, best_cost=0.247\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  28%|██▊       |14/50, best_cost=0.247\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "List of losses for all particles\n",
      "[0.4054887  0.70658024 0.64053565 5.02092012 1.09393509 0.87410698\n",
      " 0.27984301 0.40985875 8.51453535 5.16062686 0.34837901 0.85998778\n",
      " 0.8209497  0.38763998 0.91716562 5.51715079 5.89667616 0.25333516\n",
      " 0.26400256 1.05507582]\n",
      "Number of particles: 20\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  28%|██▊       |14/50, best_cost=0.239\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  30%|███       |15/50, best_cost=0.239\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "List of losses for all particles\n",
      "[0.72669419 0.98223775 1.75504824 2.54932293 0.98713666 0.28205808\n",
      " 0.23852072 1.30140814 6.87066127 2.13569314 1.1510766  0.70487434\n",
      " 0.59872745 1.5752786  1.5239206  4.85971237 6.26077705 0.34228102\n",
      " 2.69917934 1.83983227]\n",
      "Number of particles: 20\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  30%|███       |15/50, best_cost=0.239\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  32%|███▏      |16/50, best_cost=0.239\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "List of losses for all particles\n",
      "[0.35927284 1.39646835 2.17633281 1.16710815 1.10855165 0.42630272\n",
      " 0.41379902 0.44865665 0.87693808 0.91325555 1.80677875 0.49417134\n",
      " 0.99426661 2.19073039 0.50945004 1.74354485 2.03043194 0.39240453\n",
      " 1.11323434 1.5731044 ]\n",
      "Number of particles: 20\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  32%|███▏      |16/50, best_cost=0.239\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  34%|███▍      |17/50, best_cost=0.239\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "List of losses for all particles\n",
      "[1.13590721 0.99185206 1.71468628 0.82542521 1.41271154 0.57232171\n",
      " 0.48889041 0.63865182 0.8338811  0.35838658 0.66368734 4.33171553\n",
      " 1.146337   1.910767   0.40617245 0.55587338 0.39865833 0.40468291\n",
      " 0.69939027 0.70637534]\n",
      "Number of particles: 20\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  34%|███▍      |17/50, best_cost=0.167\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  36%|███▌      |18/50, best_cost=0.167\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "List of losses for all particles\n",
      "[1.30657263 0.46674075 0.79664311 0.95077722 1.88256375 0.67657854\n",
      " 0.42259271 0.72597412 0.47632822 0.90409936 0.34654491 3.90114785\n",
      " 0.84162889 1.04849296 0.42259577 0.41616865 0.77731024 0.36933237\n",
      " 0.60883555 0.16717526]\n",
      "Number of particles: 20\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  36%|███▌      |18/50, best_cost=0.151\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  38%|███▊      |19/50, best_cost=0.151\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "List of losses for all particles\n",
      "[1.00042193 0.15054399 0.72173681 0.58774985 3.04490502 0.64871561\n",
      " 0.28045452 1.84448061 0.80248025 0.95956548 0.70248282 0.71979373\n",
      " 0.35350971 0.32603473 0.90302596 0.53865466 0.50932694 0.27939351\n",
      " 0.62521551 0.47378449]\n",
      "Number of particles: 20\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  38%|███▊      |19/50, best_cost=0.151\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  40%|████      |20/50, best_cost=0.151\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "List of losses for all particles\n",
      "[0.5267732  0.15248225 1.14769693 0.54314063 2.07325013 0.37533737\n",
      " 0.18790713 1.93373442 1.13943252 0.92915799 0.87647756 0.80939813\n",
      " 0.15991882 0.38388734 0.95426425 0.71279879 0.65015572 0.33140025\n",
      " 0.85879121 0.64462058]\n",
      "Number of particles: 20\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  40%|████      |20/50, best_cost=0.151\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  42%|████▏     |21/50, best_cost=0.151\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "List of losses for all particles\n",
      "[0.23738721 0.31423698 1.21291734 0.54621523 1.7344609  0.65266075\n",
      " 0.1654562  1.0296767  1.20524605 0.59277959 0.8851773  0.70767744\n",
      " 0.62231653 0.61248845 0.75411433 0.75856341 0.43885838 0.95233435\n",
      " 0.97264507 0.66362324]\n",
      "Number of particles: 20\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  42%|████▏     |21/50, best_cost=0.133\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  44%|████▍     |22/50, best_cost=0.133\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "List of losses for all particles\n",
      "[0.31468593 0.23232649 0.76594271 0.59261759 1.62969907 1.24570037\n",
      " 0.13250394 0.2572487  1.01969518 0.19019337 0.33515959 0.48673667\n",
      " 0.97688779 0.3515106  0.78492774 1.28662254 0.35763996 1.42558521\n",
      " 0.91127228 0.42004592]\n",
      "Number of particles: 20\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  44%|████▍     |22/50, best_cost=0.133\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  46%|████▌     |23/50, best_cost=0.133\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "List of losses for all particles\n",
      "[0.36506194 0.2669002  0.7324352  0.69408935 0.87643665 1.18447777\n",
      " 0.15197065 0.81018921 0.58939019 0.52073854 0.22880528 0.22794886\n",
      " 1.08542858 0.35381954 0.9455574  1.87732037 1.36432958 1.56670385\n",
      " 0.49228954 0.23281494]\n",
      "Number of particles: 20\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  46%|████▌     |23/50, best_cost=0.133\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  48%|████▊     |24/50, best_cost=0.133\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "List of losses for all particles\n",
      "[0.33838648 0.32454902 1.07895584 0.45483279 0.567788   0.73466114\n",
      " 0.25654735 1.05185006 0.48857218 0.79721814 0.14525467 0.31535895\n",
      " 0.6073235  0.19116724 1.01599598 2.11015652 2.81876554 0.99793417\n",
      " 0.33730537 0.44514336]\n",
      "Number of particles: 20\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  48%|████▊     |24/50, best_cost=0.133\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  50%|█████     |25/50, best_cost=0.133\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "List of losses for all particles\n",
      "[0.16612017 0.4631379  1.25908584 0.154401   0.81025406 0.33663859\n",
      " 0.19445033 0.72185476 0.51529393 0.86480284 0.3187815  0.74784739\n",
      " 0.3135105  0.88333936 0.82031636 1.6353808  1.9206788  0.46589333\n",
      " 0.32018373 0.45163039]\n",
      "Number of particles: 20\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  50%|█████     |25/50, best_cost=0.133\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  52%|█████▏    |26/50, best_cost=0.133\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "List of losses for all particles\n",
      "[0.22840353 0.41243298 1.57748933 0.25688823 0.63801471 0.4683709\n",
      " 0.1815239  0.26335977 0.74325945 0.24961805 0.31484781 1.04992612\n",
      " 0.14989693 1.23700252 0.81709765 1.55583644 0.54236181 0.36992211\n",
      " 0.65657886 0.60123446]\n",
      "Number of particles: 20\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  52%|█████▏    |26/50, best_cost=0.133\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  54%|█████▍    |27/50, best_cost=0.133\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "List of losses for all particles\n",
      "[0.35022991 0.32060074 1.45459708 0.50696172 0.31930767 0.71233831\n",
      " 0.16100969 0.56066239 0.88117847 0.26696811 0.22598359 0.68715059\n",
      " 0.32733485 1.04791498 0.50181976 1.03651207 0.22021069 0.62851205\n",
      " 0.73258222 0.43652366]\n",
      "Number of particles: 20\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  54%|█████▍    |27/50, best_cost=0.133\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  56%|█████▌    |28/50, best_cost=0.133\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "List of losses for all particles\n",
      "[0.31385363 0.28931642 1.28132333 0.58116991 0.59140342 0.79523728\n",
      " 0.1495984  1.02871278 0.93061987 0.48018692 0.14795073 0.14184101\n",
      " 0.40748567 0.48576956 0.6659194  0.39387404 0.45347938 0.59940976\n",
      " 0.78515404 0.32377352]\n",
      "Number of particles: 20\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  56%|█████▌    |28/50, best_cost=0.133\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  58%|█████▊    |29/50, best_cost=0.133\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "List of losses for all particles\n",
      "[0.1458519  0.20424005 1.26768285 0.40801786 0.78130363 0.79609901\n",
      " 0.18306373 1.03087709 0.88042766 0.77831637 0.13794865 0.31468379\n",
      " 0.38331536 0.1541283  0.82663834 0.32503462 0.57460805 0.41521439\n",
      " 0.88267301 0.5752177 ]\n",
      "Number of particles: 20\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  58%|█████▊    |29/50, best_cost=0.133\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  60%|██████    |30/50, best_cost=0.133\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "List of losses for all particles\n",
      "[0.21416622 0.25087506 1.14848418 0.29662945 0.49937426 0.29731013\n",
      " 0.17674598 0.60634072 0.90616958 0.93877634 0.34160813 0.38479155\n",
      " 0.33709038 0.30545893 0.68552904 0.92170097 0.4967945  0.48265846\n",
      " 0.83648398 0.63497887]\n",
      "Number of particles: 20\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  60%|██████    |30/50, best_cost=0.133\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  62%|██████▏   |31/50, best_cost=0.133\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "List of losses for all particles\n",
      "[0.248338   0.19446152 1.1176244  0.46724137 0.32226397 1.24592466\n",
      " 0.15560571 0.34758988 0.43539552 0.69558804 0.4991891  0.32707621\n",
      " 0.22723853 0.41752182 0.67591182 1.69040007 0.41590874 0.90933123\n",
      " 0.85430041 0.49069196]\n",
      "Number of particles: 20\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  62%|██████▏   |31/50, best_cost=0.133\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  64%|██████▍   |32/50, best_cost=0.133\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "List of losses for all particles\n",
      "[0.23225818 0.16207509 1.07634415 0.57842089 0.19226208 1.40872862\n",
      " 0.15901861 0.96915306 0.38965189 0.70445449 0.40291179 0.19448016\n",
      " 0.28453449 0.41909975 0.70344374 1.9279652  0.29941215 1.03680464\n",
      " 0.77240182 0.19780131]\n",
      "Number of particles: 20\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  64%|██████▍   |32/50, best_cost=0.133\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  66%|██████▌   |33/50, best_cost=0.133\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "List of losses for all particles\n",
      "[0.13390502 0.17348863 0.86189917 0.54796671 0.25295223 1.09565025\n",
      " 0.16265021 1.04646719 0.77886785 0.55075068 0.20543075 0.27052225\n",
      " 0.39733573 0.4002882  0.63879899 1.96158765 0.51246995 0.76599557\n",
      " 0.39196564 0.36043817]\n",
      "Number of particles: 20\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  66%|██████▌   |33/50, best_cost=0.133\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  68%|██████▊   |34/50, best_cost=0.133\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "List of losses for all particles\n",
      "[0.20516734 0.17401776 1.20346729 0.39870443 0.48008249 0.61454409\n",
      " 0.14744409 0.63392657 1.25583888 0.49676701 0.16698666 0.58289101\n",
      " 0.51656819 0.30481349 0.70460248 1.68892782 0.7856825  0.53730691\n",
      " 0.45817228 0.41295054]\n",
      "Number of particles: 20\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  68%|██████▊   |34/50, best_cost=0.126\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  70%|███████   |35/50, best_cost=0.126\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "List of losses for all particles\n",
      "[0.310362   0.29326522 1.3755969  0.18767661 0.59651662 0.27675433\n",
      " 0.12623665 0.41788678 1.58391214 0.31555445 0.18095796 0.85344207\n",
      " 0.4462579  0.20073237 0.82808185 1.35611288 0.58558396 0.46131607\n",
      " 0.76994671 0.4383319 ]\n",
      "Number of particles: 20\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  70%|███████   |35/50, best_cost=0.126\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  72%|███████▏  |36/50, best_cost=0.126\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "List of losses for all particles\n",
      "[0.29951746 0.26546338 1.32001576 0.31209234 0.52138931 0.25917318\n",
      " 0.1378854  0.24822168 1.24681722 0.37640828 0.17557784 0.71837297\n",
      " 0.21987576 0.42575491 0.59896307 1.41028288 0.22556791 0.47944029\n",
      " 0.60319245 0.45958187]\n",
      "Number of particles: 20\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  72%|███████▏  |36/50, best_cost=0.126\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  74%|███████▍  |37/50, best_cost=0.126\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "List of losses for all particles\n",
      "[0.17652549 0.23789905 1.32147155 0.47879794 0.2026932  0.28515445\n",
      " 0.15010523 0.49873592 0.46333455 0.21182447 0.14614167 0.43425302\n",
      " 0.19729381 0.41518758 0.80350907 1.13186099 0.31274965 0.58749575\n",
      " 0.67865872 0.34373214]\n",
      "Number of particles: 20\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  74%|███████▍  |37/50, best_cost=0.126\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  76%|███████▌  |38/50, best_cost=0.126\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "List of losses for all particles\n",
      "[0.20630542 0.23455429 0.97440489 0.4042498  0.22021292 0.43416366\n",
      " 0.14197683 0.97359229 0.72433912 0.58969844 0.21532313 0.3166762\n",
      " 0.28682693 0.34118959 0.87642375 0.79996796 0.45385962 0.5445158\n",
      " 0.68569733 0.21969289]\n",
      "Number of particles: 20\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  76%|███████▌  |38/50, best_cost=0.122\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  78%|███████▊  |39/50, best_cost=0.122\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "List of losses for all particles\n",
      "[0.21891115 0.12159395 1.15930056 0.21228293 0.45522348 0.38756759\n",
      " 0.13106792 0.97019119 0.73936252 0.79661177 0.26131989 0.20501151\n",
      " 0.28272238 0.30995603 0.75573898 1.23706609 0.55984164 0.33894567\n",
      " 0.48775306 0.37226933]\n",
      "Number of particles: 20\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  78%|███████▊  |39/50, best_cost=0.107\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  80%|████████  |40/50, best_cost=0.107\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "List of losses for all particles\n",
      "[0.1768543  0.20213309 1.10320171 0.29786374 0.65696657 0.36110891\n",
      " 0.10733687 0.52076798 0.47288294 0.49240661 0.29443497 0.13793885\n",
      " 0.20420273 0.27436149 0.45118302 1.5140731  0.45692058 0.54276459\n",
      " 0.34585786 0.47361584]\n",
      "Number of particles: 20\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  80%|████████  |40/50, best_cost=0.107\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  82%|████████▏ |41/50, best_cost=0.107\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "List of losses for all particles\n",
      "[0.12986066 0.28872503 1.01938921 0.38350041 0.40369759 0.47762954\n",
      " 0.1296921  0.22287134 0.24710542 0.35040627 0.19010385 0.20767481\n",
      " 0.13417711 0.23719145 0.35598508 1.42785137 0.35198223 0.96208204\n",
      " 0.33558309 0.44159188]\n",
      "Number of particles: 20\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  82%|████████▏ |41/50, best_cost=0.106\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  84%|████████▍ |42/50, best_cost=0.106\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "List of losses for all particles\n",
      "[0.13018536 0.2313979  0.9744945  0.39228339 0.26363498 0.6640244\n",
      " 0.15355703 0.21539243 0.45423204 0.3606064  0.18291138 0.24275463\n",
      " 0.10558797 0.23126285 0.47681055 1.40729938 0.19735472 0.97801018\n",
      " 0.38030209 0.18128002]\n",
      "Number of particles: 20\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  84%|████████▍ |42/50, best_cost=0.106\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  86%|████████▌ |43/50, best_cost=0.106\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "List of losses for all particles\n",
      "[0.11768179 0.11350446 0.71226798 0.34806435 0.64616374 0.62689513\n",
      " 0.17409284 0.2002265  0.60709267 0.4794755  0.27054021 0.21912931\n",
      " 0.18396857 0.3570777  0.54230884 1.25251339 0.36995946 0.42260492\n",
      " 0.37551871 0.36119362]\n",
      "Number of particles: 20\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  86%|████████▌ |43/50, best_cost=0.106\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  88%|████████▊ |44/50, best_cost=0.106\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "List of losses for all particles\n",
      "[0.10744191 0.1683692  0.66285313 0.2827577  1.0670568  0.65427593\n",
      " 0.14732347 0.2167775  0.50773551 0.39245033 0.35427897 0.21764632\n",
      " 0.32862451 0.29793872 0.43641318 0.69163746 0.60502981 0.31485916\n",
      " 0.56435738 0.65806146]\n",
      "Number of particles: 20\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  88%|████████▊ |44/50, best_cost=0.106\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  90%|█████████ |45/50, best_cost=0.106\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "List of losses for all particles\n",
      "[0.11078883 0.28151244 0.89253698 0.15937265 1.00259793 0.59029476\n",
      " 0.10673077 0.15877699 0.40880188 0.18295417 0.35695945 0.18983479\n",
      " 0.35365536 0.51515987 0.22720732 0.58305121 0.44054095 0.5448268\n",
      " 0.60583614 0.52903812]\n",
      "Number of particles: 20\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  90%|█████████ |45/50, best_cost=0.106\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  92%|█████████▏|46/50, best_cost=0.106\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "List of losses for all particles\n",
      "[0.11439119 0.26851842 1.0834162  0.12608944 0.58576192 0.75500542\n",
      " 0.13460224 0.33237554 0.42478539 0.15501348 0.21516639 0.11838877\n",
      " 0.26218995 0.46805066 0.37279569 0.70813822 0.19305259 0.57104845\n",
      " 0.53420598 0.14162887]\n",
      "Number of particles: 20\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  92%|█████████▏|46/50, best_cost=0.0988\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  94%|█████████▍|47/50, best_cost=0.0988\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "List of losses for all particles\n",
      "[0.12748475 0.1456869  1.31686601 0.09881684 0.32567248 0.76457204\n",
      " 0.18017705 0.2858754  0.49323046 0.17193351 0.14172547 0.10091055\n",
      " 0.13325161 0.29244027 0.61889587 1.2489189  0.24879329 0.4751286\n",
      " 0.29665353 0.25616911]\n",
      "Number of particles: 20\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  94%|█████████▍|47/50, best_cost=0.0988\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  96%|█████████▌|48/50, best_cost=0.0988\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "List of losses for all particles\n",
      "[0.13245571 0.12167562 1.53478457 0.2083861  0.7123124  0.38650569\n",
      " 0.18771763 0.20751858 0.58281693 0.27469204 0.20256069 0.17974856\n",
      " 0.11809953 0.20650115 0.62423766 1.91102214 0.22268355 0.39065599\n",
      " 0.24693691 0.36947933]\n",
      "Number of particles: 20\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  96%|█████████▌|48/50, best_cost=0.0988\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  98%|█████████▊|49/50, best_cost=0.0988\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "List of losses for all particles\n",
      "[0.14673934 0.12764288 1.35834483 0.3326591  1.05784575 0.22975338\n",
      " 0.11901528 0.18884001 0.45992764 0.31472279 0.3440851  0.23929362\n",
      " 0.16849162 0.20971482 0.34731384 1.47657384 0.14920415 0.33601609\n",
      " 0.35753619 0.35441269]\n",
      "Number of particles: 20\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best:  98%|█████████▊|49/50, best_cost=0.0881\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "pyswarms.single.global_best: 100%|██████████|50/50, best_cost=0.0881\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A2019-03-02 15:05:09,080 - pyswarms.single.global_best - INFO - Optimization finished | best cost: 0.08814329879892738, best pos: [-0.02514773  0.17646692 -0.37365008 ... -0.2593118  -0.95572563\n",
      "  0.40355033]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "List of losses for all particles\n",
      "[0.20697627 0.24063233 1.30655824 0.31225573 1.00152663 0.25391254\n",
      " 0.09240415 0.1716728  0.22891094 0.15417872 0.4198473  0.24553019\n",
      " 0.20329018 0.24411249 0.20148165 1.06295699 0.0881433  0.78218947\n",
      " 0.37424598 0.2134111 ]\n"
     ]
    }
   ],
   "source": [
    "options = {'c1': 0.5, 'c2': 0.3, 'w':0.9}\n",
    "\n",
    "layer_1 = modelGA.get_layer('layer_1').get_weights()\n",
    "layer_2 = modelGA.get_layer('layer_2').get_weights()\n",
    "dimensions = (layer_1[0].shape[0] * layer_1[0].shape[1]) + (layer_1[0].shape[0] * layer_1[0].shape[1]) + (layer_1[1].shape[0] * layer_1[1].shape[1]) + (layer_2[0].shape[0] * layer_2[0].shape[1]) +(layer_2[1].shape[0])\n",
    "\n",
    "print(\"Number of params in Neural Network: %d\" % dimensions)\n",
    "\n",
    "optimizer = ps.single.GlobalBestPSO(n_particles=20, dimensions=dimensions, options=options)\n",
    "\n",
    "cost, pos = optimizer.optimize(f, iters=50)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 431,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_cost_history(cost_history=optimizer.cost_history)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Creating the Neural Network with the final optimized params using GA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 446,
   "metadata": {},
   "outputs": [],
   "source": [
    "layer_1 = modelGA.get_layer('layer_1').get_weights()\n",
    "layer_2 = modelGA.get_layer('layer_2').get_weights()\n",
    "\n",
    "layer1_W1_shape = (layer_1[0].shape[0] * layer_1[0].shape[1])\n",
    "layer1_W2_shape = layer1_W1_shape + (layer_1[1].shape[0] * layer_1[1].shape[1])\n",
    "layer1_B_shape = layer1_W2_shape + (layer_1[2].shape[0])\n",
    "layer2_W_shape = layer1_B_shape + (layer_2[0].shape[0] * layer_2[0].shape[1])\n",
    "layer2_B_shape = layer2_W_shape + (layer_2[1].shape[0])\n",
    "\n",
    "layer1_W1 = pos[0:layer1_W1_shape].reshape(layer_1[0].shape)\n",
    "layer1_W2 = pos[layer1_W1_shape:layer1_W2_shape].reshape(layer_1[1].shape)\n",
    "layer1_B = pos[layer1_W2_shape:layer1_B_shape].reshape(layer_1[2].shape)\n",
    "layer2_W = pos[layer1_B_shape:layer2_W_shape].reshape(layer_2[0].shape)\n",
    "layer2_B = pos[layer2_W_shape:layer2_B_shape].reshape(layer_2[1].shape)\n",
    "\n",
    "modelGA.get_layer('layer_1').set_weights([layer1_W1, layer1_W2, layer1_B])\n",
    "modelGA.get_layer('layer_2').set_weights([layer2_W, layer2_B])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Checking the performance of NN using GA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 447,
   "metadata": {},
   "outputs": [],
   "source": [
    "predictY = modelGA.predict(testX)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 448,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "time = np.arange(predictY.shape[0])\n",
    "\n",
    "plt.plot(time, predictY, testY)\n",
    "plt.xlabel('time')\n",
    "plt.ylabel('amplitude')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 449,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 0s 133us/step\n"
     ]
    }
   ],
   "source": [
    "lossNN_GA = modelGA.evaluate(testX, testY)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 450,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The loss output for test dataset is: 0.007762\n"
     ]
    }
   ],
   "source": [
    "print('The loss output for test dataset is: %f' % lossNN_GA[0])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Conclusion"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 455,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "We see that the loss output for test dataset using Back Propagation is: 0.067104, while the loss output using GA is: 0.007762\n"
     ]
    }
   ],
   "source": [
    "print('We see that the loss output for test dataset using Back Propagation is: %f, while the loss output using GA is: %f' % (lossNN[0], lossNN_GA[0]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### References\n",
    "- https://pyswarms.readthedocs.io/en/latest/examples/visualization.html\n",
    "- https://keras.io/getting-started/sequential-model-guide/"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}