examples/10_minutes_to_pypolars.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Looking in indexes: https://test.pypi.org/simple/\n",
      "Requirement already up-to-date: polars in /opt/miniconda3/lib/python3.7/site-packages (0.0.2)\n",
      "Requirement already satisfied, skipping upgrade: numpy in /opt/miniconda3/lib/python3.7/site-packages (from polars) (1.19.0)\n",
      "Requirement already satisfied: tqdm in /opt/miniconda3/lib/python3.7/site-packages (4.32.1)\n",
      "Requirement already satisfied: pandas in /opt/miniconda3/lib/python3.7/site-packages (1.0.5)\n",
      "Requirement already satisfied: matplotlib in /opt/miniconda3/lib/python3.7/site-packages (3.1.3)\n",
      "Requirement already satisfied: numpy>=1.13.3 in /opt/miniconda3/lib/python3.7/site-packages (from pandas) (1.19.0)\n",
      "Requirement already satisfied: python-dateutil>=2.6.1 in /opt/miniconda3/lib/python3.7/site-packages (from pandas) (2.8.1)\n",
      "Requirement already satisfied: pytz>=2017.2 in /opt/miniconda3/lib/python3.7/site-packages (from pandas) (2020.1)\n",
      "Requirement already satisfied: pyparsing!=2.0.4,!=2.1.2,!=2.1.6,>=2.0.1 in /opt/miniconda3/lib/python3.7/site-packages (from matplotlib) (2.4.7)\n",
      "Requirement already satisfied: cycler>=0.10 in /opt/miniconda3/lib/python3.7/site-packages (from matplotlib) (0.10.0)\n",
      "Requirement already satisfied: kiwisolver>=1.0.1 in /opt/miniconda3/lib/python3.7/site-packages (from matplotlib) (1.2.0)\n",
      "Requirement already satisfied: six>=1.5 in /opt/miniconda3/lib/python3.7/site-packages (from python-dateutil>=2.6.1->pandas) (1.12.0)\n",
      "Requirement already satisfied: pandas in /opt/miniconda3/lib/python3.7/site-packages (1.0.5)\n",
      "Requirement already satisfied: pytz>=2017.2 in /opt/miniconda3/lib/python3.7/site-packages (from pandas) (2020.1)\n",
      "Requirement already satisfied: numpy>=1.13.3 in /opt/miniconda3/lib/python3.7/site-packages (from pandas) (1.19.0)\n",
      "Requirement already satisfied: python-dateutil>=2.6.1 in /opt/miniconda3/lib/python3.7/site-packages (from pandas) (2.8.1)\n",
      "Requirement already satisfied: six>=1.5 in /opt/miniconda3/lib/python3.7/site-packages (from python-dateutil>=2.6.1->pandas) (1.12.0)\n"
     ]
    }
   ],
   "source": [
    "!pip install --upgrade polars\n",
    "!pip install tqdm pandas matplotlib"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 10 minutes to py-polars\n",
    "This as short introduction to Polars to get you started with the basic concepts of data wrangling. It is very much influenced by [10 minutes to pandas](https://pandas.pydata.org/pandas-docs/stable/user_guide/10min.html).\n",
    "\n",
    "py-polars are the python bindings to Polars. It currently supports only a subset of the datatypes and operations supported by Polars. \n",
    "However it should be enough to give your slow pipelines a boost."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import polars as pl\n",
    "import numpy as np\n",
    "np.random.seed(1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Object creation\n",
    "\n",
    "Creating a `Series` by passing a list or array of values."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Series: 'a' [i64]\n",
       "[\n",
       "\t1\n",
       "\t2\n",
       "\t3\n",
       "]"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pl.Series(\"a\", [1, 2, 3])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "A `Series` can also have nullable values."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Series: 'with nullable values' [i64]\n",
       "[\n",
       "\t1\n",
       "\tnull\n",
       "\t3\n",
       "]"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "s = pl.Series(\"with nullable values\", [1, None, 3], nullable=True)\n",
    "s"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Series have a data type and can be casted"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'polars.datatypes.Int64'>\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "Series: 'with nullable values' [f32]\n",
       "[\n",
       "\t1\n",
       "\tnull\n",
       "\t3\n",
       "]"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "print(s.dtype)\n",
    "s.cast_f32()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "A `DataFrame` can be created by passing a dictionary with keys as column names and \n",
    "list values."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "+-------+-----+-----+\n",
       "| foo   | bar | ham |\n",
       "| ---   | --- | --- |\n",
       "| f64   | i64 | str |\n",
       "+=======+=====+=====+\n",
       "| 0.417 | 0   | \"h\" |\n",
       "+-------+-----+-----+\n",
       "| 0.72  | 1   | \"h\" |\n",
       "+-------+-----+-----+\n",
       "| 0e0   | 2   | \"h\" |\n",
       "+-------+-----+-----+"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pl.DataFrame({\n",
    "    \"foo\": np.random.rand(10),\n",
    "    \"bar\": np.arange(10),\n",
    "    \"ham\": [\"h\"] * 3 + [\"a\"] * 3 + [\"m\"] * 4\n",
    "})\n",
    "df.head(3)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The columns of the result `DataFrame` have different types and names"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[<class 'polars.datatypes.Float64'>, <class 'polars.datatypes.Int64'>, <class 'polars.datatypes.Utf8'>]\n",
      "['foo', 'bar', 'ham']\n"
     ]
    }
   ],
   "source": [
    "print(df.dtypes)\n",
    "print(df.columns)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Viewing data\n",
    "\n",
    "We can view the top and bottom rows of a `DataFrame`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "+-------+-----+-----+\n",
       "| foo   | bar | ham |\n",
       "| ---   | --- | --- |\n",
       "| f64   | i64 | str |\n",
       "+=======+=====+=====+\n",
       "| 0.417 | 0   | \"h\" |\n",
       "+-------+-----+-----+\n",
       "| 0.72  | 1   | \"h\" |\n",
       "+-------+-----+-----+\n",
       "| 0e0   | 2   | \"h\" |\n",
       "+-------+-----+-----+"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head(3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "+-------+-----+-----+\n",
       "| foo   | bar | ham |\n",
       "| ---   | --- | --- |\n",
       "| f64   | i64 | str |\n",
       "+=======+=====+=====+\n",
       "| 0.346 | 7   | \"m\" |\n",
       "+-------+-----+-----+\n",
       "| 0.397 | 8   | \"m\" |\n",
       "+-------+-----+-----+\n",
       "| 0.539 | 9   | \"m\" |\n",
       "+-------+-----+-----+"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.tail(3)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can sort by column."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "+-------+-----+-----+\n",
       "| foo   | bar | ham |\n",
       "| ---   | --- | --- |\n",
       "| f64   | i64 | str |\n",
       "+=======+=====+=====+\n",
       "| 0.72  | 1   | \"h\" |\n",
       "+-------+-----+-----+\n",
       "| 0.539 | 9   | \"m\" |\n",
       "+-------+-----+-----+\n",
       "| 0.417 | 0   | \"h\" |\n",
       "+-------+-----+-----+\n",
       "| 0.397 | 8   | \"m\" |\n",
       "+-------+-----+-----+\n",
       "| 0.346 | 7   | \"m\" |\n",
       "+-------+-----+-----+"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.sort(\"foo\", reverse=True).head(5)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Selection\n",
    "We can select a single column, which returns a Series."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Series: 'foo' [f64]\n",
       "[\n",
       "\t0.417\n",
       "\t0.72\n",
       "\t0e0\n",
       "]"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[\"foo\"].head(3)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Or select a column by index"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Series: 'foo' [f64]\n",
       "[\n",
       "\t0.417\n",
       "\t0.72\n",
       "\t0e0\n",
       "\t0.302\n",
       "\t0.147\n",
       "\t0.092\n",
       "\t0.186\n",
       "\t0.346\n",
       "\t0.397\n",
       "\t0.539\n",
       "]"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[0]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "When we select in two dimensions, we select by row, column order.\n",
    "Here we slice until the third row of the first column."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Series: 'foo' [f64]\n",
       "[\n",
       "\t0.417\n",
       "\t0.72\n",
       "\t0e0\n",
       "]"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[:3, 0]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Or we can slice the whole `DataFrame` into a smaller sub `DataFrame`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "+-------+-----+-----+\n",
       "| foo   | bar | ham |\n",
       "| ---   | --- | --- |\n",
       "| f64   | i64 | str |\n",
       "+=======+=====+=====+\n",
       "| 0.417 | 0   | \"h\" |\n",
       "+-------+-----+-----+\n",
       "| 0.72  | 1   | \"h\" |\n",
       "+-------+-----+-----+\n",
       "| 0e0   | 2   | \"h\" |\n",
       "+-------+-----+-----+\n",
       "| 0.302 | 3   | \"a\" |\n",
       "+-------+-----+-----+"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[:4]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Or we slice both rows and columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "+-----+-----+\n",
       "| bar | ham |\n",
       "| --- | --- |\n",
       "| i64 | str |\n",
       "+=====+=====+\n",
       "| 3   | \"a\" |\n",
       "+-----+-----+\n",
       "| 4   | \"a\" |\n",
       "+-----+-----+"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[3:5,1:]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Boolean indexing\n",
    "Boolean indexes can be used to filter data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "+-------+-----+-----+\n",
       "| foo   | bar | ham |\n",
       "| ---   | --- | --- |\n",
       "| f64   | i64 | str |\n",
       "+=======+=====+=====+\n",
       "| 0.72  | 1   | \"h\" |\n",
       "+-------+-----+-----+\n",
       "| 0.539 | 9   | \"m\" |\n",
       "+-------+-----+-----+"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[df[\"foo\"] > 0.5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "+-------+-----+-----+\n",
       "| foo   | bar | ham |\n",
       "| ---   | --- | --- |\n",
       "| f64   | i64 | str |\n",
       "+=======+=====+=====+\n",
       "| 0.302 | 3   | \"a\" |\n",
       "+-------+-----+-----+\n",
       "| 0.147 | 4   | \"a\" |\n",
       "+-------+-----+-----+\n",
       "| 0.092 | 5   | \"a\" |\n",
       "+-------+-----+-----+"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[df[\"ham\"] == \"a\"]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Setting\n",
    "Adding a new column to the `DataFrame` can be done with `hstack`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "+-------+-----+-----+-----+\n",
       "| foo   | bar | ham | new |\n",
       "| ---   | --- | --- | --- |\n",
       "| f64   | i64 | str | i64 |\n",
       "+=======+=====+=====+=====+\n",
       "| 0.417 | 0   | \"h\" | 0   |\n",
       "+-------+-----+-----+-----+\n",
       "| 0.72  | 1   | \"h\" | 1   |\n",
       "+-------+-----+-----+-----+\n",
       "| 0e0   | 2   | \"h\" | 2   |\n",
       "+-------+-----+-----+-----+\n",
       "| 0.302 | 3   | \"a\" | 3   |\n",
       "+-------+-----+-----+-----+\n",
       "| 0.147 | 4   | \"a\" | 4   |\n",
       "+-------+-----+-----+-----+\n",
       "| 0.092 | 5   | \"a\" | 5   |\n",
       "+-------+-----+-----+-----+\n",
       "| 0.186 | 6   | \"m\" | 6   |\n",
       "+-------+-----+-----+-----+\n",
       "| 0.346 | 7   | \"m\" | 7   |\n",
       "+-------+-----+-----+-----+\n",
       "| 0.397 | 8   | \"m\" | 8   |\n",
       "+-------+-----+-----+-----+\n",
       "| 0.539 | 9   | \"m\" | 9   |\n",
       "+-------+-----+-----+-----+"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[\"new\"] = np.arange(10)\n",
    "df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can also define the column location by index"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "+---------+-----+-----+-----+\n",
       "| new_foo | bar | ham | new |\n",
       "| ---     | --- | --- | --- |\n",
       "| f64     | i64 | str | i64 |\n",
       "+=========+=====+=====+=====+\n",
       "| 0.419   | 0   | \"h\" | 0   |\n",
       "+---------+-----+-----+-----+\n",
       "| 0.685   | 1   | \"h\" | 1   |\n",
       "+---------+-----+-----+-----+\n",
       "| 0.204   | 2   | \"h\" | 2   |\n",
       "+---------+-----+-----+-----+"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[0] = pl.Series(\"new_foo\", np.random.rand(10))\n",
    "df.head(3)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Or use a boolean mask to assign new values.\n",
    "\n",
    "_Note that every mutable assignment alocates new memory. So isn't actually mutable with regard to the actual memory. This is a performance trade off. Due to the immutable memory, slices, clones, subsets of `Series`/`DataFrames` are zero copy. If you need to mutate a lot of values, it's faster to do this in numpy allocate a new `Series`._"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "# selection order is row column\n",
    "df[df[\"new_foo\"] > 0.5, \"new_foo\"] = 1"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "we can also define the mutation location by passing an array of indexes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Series: 'ham' [str]\n",
       "[\n",
       "\t\"h\"\n",
       "\t\"c\"\n",
       "\t\"c\"\n",
       "\t\"c\"\n",
       "\t\"a\"\n",
       "\t\"a\"\n",
       "\t\"m\"\n",
       "\t\"m\"\n",
       "\t\"m\"\n",
       "\t\"m\"\n",
       "]"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "s = df[\"ham\"]\n",
    "s[[1, 2, 3]] = \"c\"\n",
    "s"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Concat\n",
    "Polars provide methods to cobine multiple `DataFrames` and `Series`.\n",
    "We can concatenate a `DataFrame` with `hstack`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "+---------+-----+-----+-----+---------+-----+-----+-----+\n",
       "| new_foo | bar | ham | new | new_foo | bar | ham | new |\n",
       "| ---     | --- | --- | --- | ---     | --- | --- | --- |\n",
       "| f64     | i64 | str | i64 | f64     | i64 | str | i64 |\n",
       "+=========+=====+=====+=====+=========+=====+=====+=====+\n",
       "| 0.419   | 0   | \"h\" | 0   | 0.419   | 0   | \"h\" | 0   |\n",
       "+---------+-----+-----+-----+---------+-----+-----+-----+\n",
       "| 1       | 1   | \"h\" | 1   | 1       | 1   | \"h\" | 1   |\n",
       "+---------+-----+-----+-----+---------+-----+-----+-----+\n",
       "| 0.204   | 2   | \"h\" | 2   | 0.204   | 2   | \"h\" | 2   |\n",
       "+---------+-----+-----+-----+---------+-----+-----+-----+\n",
       "| 1       | 3   | \"a\" | 3   | 1       | 3   | \"a\" | 3   |\n",
       "+---------+-----+-----+-----+---------+-----+-----+-----+\n",
       "| 0.027   | 4   | \"a\" | 4   | 0.027   | 4   | \"a\" | 4   |\n",
       "+---------+-----+-----+-----+---------+-----+-----+-----+"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# clones are super cheap!\n",
    "df1 = df.clone()\n",
    "df1.hstack(df.get_columns())\n",
    "df1.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Or append rows from another `DataFrame`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "10 20\n"
     ]
    }
   ],
   "source": [
    "df1 = df.clone()\n",
    "df1.vstack(df)\n",
    "print(df.height, df1.height)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Joins\n",
    "SQL-styel joins."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "+-------+------+------+\n",
       "| key   | lval | rval |\n",
       "| ---   | ---  | ---  |\n",
       "| str   | i64  | i64  |\n",
       "+=======+======+======+\n",
       "| \"foo\" | 1    | 4    |\n",
       "+-------+------+------+\n",
       "| \"foo\" | 2    | 4    |\n",
       "+-------+------+------+\n",
       "| \"foo\" | 1    | 5    |\n",
       "+-------+------+------+\n",
       "| \"foo\" | 2    | 5    |\n",
       "+-------+------+------+"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "left = pl.DataFrame({'key': ['foo', 'foo'], 'lval': [1, 2]})\n",
    "right = pl.DataFrame({'key': ['foo', 'foo'], 'rval': [4, 5]})\n",
    "\n",
    "left.join(right, left_on=\"key\", right_on=\"key\", how=\"inner\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Another example that can be given is:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "+-------+------+------+\n",
       "| key   | lval | rval |\n",
       "| ---   | ---  | ---  |\n",
       "| str   | i64  | i64  |\n",
       "+=======+======+======+\n",
       "| \"foo\" | 1    | 4    |\n",
       "+-------+------+------+\n",
       "| \"bar\" | 2    | 5    |\n",
       "+-------+------+------+"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "left = pl.DataFrame({'key': ['foo', 'bar'], 'lval': [1, 2]})\n",
    "right = pl.DataFrame({'key': ['foo', 'bar'], 'rval': [4, 5]})\n",
    "\n",
    "left.join(right, left_on=\"key\", right_on=\"key\", how=\"inner\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Grouping\n",
    "\n",
    "By \"group by\" we are referring to a process involving one or more of the following steps:\n",
    "* **Splitting** the data into groups based on some criteria\n",
    "* **Applying** a function to each group independently\n",
    "* **Combining** the results into a data structure"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "+-------+---------+----------+----------+\n",
       "| A     | B       | C        | D        |\n",
       "| ---   | ---     | ---      | ---      |\n",
       "| str   | str     | f64      | f64      |\n",
       "+=======+=========+==========+==========+\n",
       "| \"foo\" | \"one\"   | 1.134    | 0.902    |\n",
       "+-------+---------+----------+----------+\n",
       "| \"bar\" | \"one\"   | -1.1e0   | 0.502    |\n",
       "+-------+---------+----------+----------+\n",
       "| \"foo\" | \"two\"   | -1.72e-1 | 0.901    |\n",
       "+-------+---------+----------+----------+\n",
       "| \"bar\" | \"three\" | -8.78e-1 | -6.84e-1 |\n",
       "+-------+---------+----------+----------+\n",
       "| \"foo\" | \"two\"   | 0.042    | -1.23e-1 |\n",
       "+-------+---------+----------+----------+\n",
       "| \"bar\" | \"two\"   | 0.583    | -9.36e-1 |\n",
       "+-------+---------+----------+----------+\n",
       "| \"foo\" | \"one\"   | -1.101e0 | -2.68e-1 |\n",
       "+-------+---------+----------+----------+\n",
       "| \"foo\" | \"three\" | 1.145    | 0.53     |\n",
       "+-------+---------+----------+----------+"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pl.DataFrame({'A': ['foo', 'bar', 'foo', 'bar',\n",
    "                         'foo', 'bar', 'foo', 'foo'],\n",
    "                   'B': ['one', 'one', 'two', 'three',\n",
    "                         'two', 'two', 'one', 'three'],\n",
    "                   'C': np.random.randn(8),\n",
    "                   'D': np.random.randn(8)})\n",
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "+-------+----------+----------+\n",
       "| A     | C_sum    | D_sum    |\n",
       "| ---   | ---      | ---      |\n",
       "| str   | f64      | f64      |\n",
       "+=======+==========+==========+\n",
       "| \"foo\" | 1.048    | 1.942    |\n",
       "+-------+----------+----------+\n",
       "| \"bar\" | -1.395e0 | -1.117e0 |\n",
       "+-------+----------+----------+"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.groupby(\"A\").select([\"C\", \"D\"]).sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "+-------+---------+---------+---------+----------+----------+\n",
       "| A     | B       | A_first | B_first | C_first  | D_first  |\n",
       "| ---   | ---     | ---     | ---     | ---      | ---      |\n",
       "| str   | str     | str     | str     | f64      | f64      |\n",
       "+=======+=========+=========+=========+==========+==========+\n",
       "| \"bar\" | \"three\" | \"bar\"   | \"three\" | -8.78e-1 | -6.84e-1 |\n",
       "+-------+---------+---------+---------+----------+----------+\n",
       "| \"foo\" | \"one\"   | \"foo\"   | \"one\"   | 1.134    | 0.902    |\n",
       "+-------+---------+---------+---------+----------+----------+\n",
       "| \"bar\" | \"two\"   | \"bar\"   | \"two\"   | 0.583    | -9.36e-1 |\n",
       "+-------+---------+---------+---------+----------+----------+\n",
       "| \"foo\" | \"three\" | \"foo\"   | \"three\" | 1.145    | 0.53     |\n",
       "+-------+---------+---------+---------+----------+----------+\n",
       "| \"bar\" | \"one\"   | \"bar\"   | \"one\"   | -1.1e0   | 0.502    |\n",
       "+-------+---------+---------+---------+----------+----------+\n",
       "| \"foo\" | \"two\"   | \"foo\"   | \"two\"   | -1.72e-1 | 0.901    |\n",
       "+-------+---------+---------+---------+----------+----------+"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.groupby([\"A\", \"B\"]).select_all().first()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "+-------+-----------------+\n",
       "| A     | C_quantile_0.20 |\n",
       "| ---   | ---             |\n",
       "| str   | f64             |\n",
       "+=======+=================+\n",
       "| \"foo\" | -1.101e0        |\n",
       "+-------+-----------------+\n",
       "| \"bar\" | -1.1e0          |\n",
       "+-------+-----------------+"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.groupby(\"A\").select(\"C\").quantile(0.2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Pivot tables\n",
    "\n",
    "Pivots create a summary table by a applying a groupby and defining a pivot column and values to aggregate."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pl.DataFrame({'A': ['one', 'one', 'two', 'three'] * 3,\n",
    "                   'B': ['A', 'B', 'C'] * 4,\n",
    "                   'C': ['foo', 'foo', 'foo', 'bar', 'bar', 'bar'] * 2,\n",
    "                   'D': np.random.randn(12),\n",
    "                   'E': np.random.randn(12)})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "+---------+-------+----------+\n",
       "| A       | foo   | bar      |\n",
       "| ---     | ---   | ---      |\n",
       "| str     | f64   | f64      |\n",
       "+=========+=======+==========+\n",
       "| \"one\"   | 1.245 | 1.939    |\n",
       "+---------+-------+----------+\n",
       "| \"three\" | 0.617 | -9.86e-1 |\n",
       "+---------+-------+----------+\n",
       "| \"two\"   | 0.171 | -1.143e0 |\n",
       "+---------+-------+----------+"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.groupby(\"A\").pivot(pivot_column=\"C\", values_column=\"E\").sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "+---------+-----+----------+----------+\n",
       "| A       | B   | foo      | bar      |\n",
       "| ---     | --- | ---      | ---      |\n",
       "| str     | str | f64      | f64      |\n",
       "+=========+=====+==========+==========+\n",
       "| \"three\" | \"A\" | null     | -6.37e-1 |\n",
       "+---------+-----+----------+----------+\n",
       "| \"one\"   | \"C\" | 0.3      | 2.1      |\n",
       "+---------+-----+----------+----------+\n",
       "| \"two\"   | \"A\" | 0.12     | null     |\n",
       "+---------+-----+----------+----------+\n",
       "| \"one\"   | \"A\" | -7.47e-1 | -3.52e-1 |\n",
       "+---------+-----+----------+----------+\n",
       "| \"two\"   | \"C\" | 0.051    | null     |\n",
       "+---------+-----+----------+----------+\n",
       "| \"three\" | \"B\" | 0.617    | null     |\n",
       "+---------+-----+----------+----------+\n",
       "| \"one\"   | \"B\" | 1.692    | 0.191    |\n",
       "+---------+-----+----------+----------+\n",
       "| \"two\"   | \"B\" | null     | -1.143e0 |\n",
       "+---------+-----+----------+----------+\n",
       "| \"three\" | \"C\" | null     | -3.49e-1 |\n",
       "+---------+-----+----------+----------+"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pivotted = df.groupby([\"A\", \"B\"]).pivot(pivot_column=\"C\", values_column=\"E\").sum()\n",
    "pivotted"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Numpy interops\n",
    "\n",
    "Polars has zero cost interaction with numpy's [ufunc](https://numpy.org/doc/stable/reference/ufuncs.html) functionality. This means that if a function/ method isn't supported by Polars, we can use numpy's without any overhead. Numpy will write the output to Polars/ arrow memory, and the null bitmask will keep null information."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Series: 'bar' [f64]\n",
       "[\n",
       "\t0.804\n",
       "\t-5.05e-1\n",
       "\tnull\n",
       "\t0.939\n",
       "\tnull\n",
       "\tnull\n",
       "\t0.982\n",
       "\t0.415\n",
       "\t0.94\n",
       "]"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "s = pivotted[\"bar\"]\n",
    "np.cos(s)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Series: 'bar' [f64]\n",
       "[\n",
       "\t0.529\n",
       "\t8.168\n",
       "\tnull\n",
       "\t0.703\n",
       "\tnull\n",
       "\tnull\n",
       "\t1.21\n",
       "\t0.319\n",
       "\t0.705\n",
       "]"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.exp(s)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Clones\n",
    "It was already mentioned previously, but clones are super cheap. This is due to the fact that the underlying memory backed by Polars is immutable. Slices and clones can be made with the guarantee that they will never be modified.\n",
    "\n",
    "Below we observe that cloning an array of 1e6 elements is almost 850x faster. The cost of clone a Polars Series is also constant and doesn't increase with memory size. Ideal for writing pure funcitons. The cost of cloning a DataFrame 10x the size is also very small."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "a = np.arange(int(1e6))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "692 µs ± 41.5 µs per loop (mean ± std. dev. of 7 runs, 1000 loops each)\n"
     ]
    }
   ],
   "source": [
    "%%timeit\n",
    "np.copy(a)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "s = pl.Series(\"a\", a)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "768 ns ± 45.2 ns per loop (mean ± std. dev. of 7 runs, 1000000 loops each)\n"
     ]
    }
   ],
   "source": [
    "%%timeit\n",
    "s.clone()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1000000, 10)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+\n",
       "| a_0 | a_1 | a_2 | a_3 | a_4 | a_5 | a_6 | a_7 | a_8 | a_9 |\n",
       "| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |\n",
       "| i64 | i64 | i64 | i64 | i64 | i64 | i64 | i64 | i64 | i64 |\n",
       "+=====+=====+=====+=====+=====+=====+=====+=====+=====+=====+\n",
       "| 0   | 0   | 0   | 0   | 0   | 0   | 0   | 0   | 0   | 0   |\n",
       "+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+\n",
       "| 1   | 1   | 1   | 1   | 1   | 1   | 1   | 1   | 1   | 1   |\n",
       "+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+\n",
       "| 2   | 2   | 2   | 2   | 2   | 2   | 2   | 2   | 2   | 2   |\n",
       "+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pl.DataFrame({f\"a_{i}\": s.clone() for i in range(10)})\n",
    "print(df.shape)\n",
    "df.head(3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1.68 µs ± 21 ns per loop (mean ± std. dev. of 7 runs, 1000000 loops each)\n"
     ]
    }
   ],
   "source": [
    "%%timeit\n",
    "df.clone()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Performance\n",
    "Let's check some performances with differences with Pandas. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import tqdm\n",
    "from pandas.util.testing import rands\n",
    "import time\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "def create_join_dfs(N = 10_000):\n",
    "    left_pivot = int(0.8 * N)\n",
    "    right_pivot = N - left_pivot\n",
    "\n",
    "    indices = np.array([rands(10) for _ in range(N)], dtype=\"O\")\n",
    "    indices2 = np.array([rands(10) for _ in range(N)], dtype=\"O\")\n",
    "    key = np.tile(indices[:left_pivot], 10)\n",
    "    key2 = np.tile(indices2[:left_pivot], 10)\n",
    "\n",
    "    left = pd.DataFrame({\"key\": key, \"key2\": key2, \"value\": np.random.randn(len(key))})\n",
    "    right = pd.DataFrame(\n",
    "        {\"key\": indices[right_pivot:], \"key2\": indices2[right_pivot:], \"value2\": np.random.randn(left_pivot)}\n",
    "    )\n",
    "    return left, right\n",
    "    \n",
    "left, right = create_join_dfs()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.0002286"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def time_lambda(f: \"fn() -> ()\") -> float:\n",
    "    \"\"\"\n",
    "    eval time in ms\n",
    "    \"\"\"\n",
    "    t0 = time.time_ns() \n",
    "    for _ in range(10):\n",
    "        f()\n",
    "    return (time.time_ns() - t0) / 10 / 1e6\n",
    "\n",
    "time_lambda(lambda : 1)\n",
    "        "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 9/9 [13:34<00:00, 203.94s/it]\n"
     ]
    }
   ],
   "source": [
    "left_pl = []\n",
    "left_pd = []\n",
    "inner_pd = []\n",
    "inner_pl = []\n",
    "outer_pd = []\n",
    "outer_pl = []\n",
    "par_inner = []\n",
    "par_left = []\n",
    "n_proxy = []\n",
    "\n",
    "for N in tqdm.tqdm([10, 30, 40, 50, 70, 100, 500, 1000, 2000]):\n",
    "    N *= 1000\n",
    "    left, right = create_join_dfs(N)\n",
    "    f_left_pd = lambda: left.merge(right, on=\"key\", how=\"left\")\n",
    "    f_inner_pd = lambda: left.merge(right, on=\"key\", how=\"inner\")\n",
    "    f_outer_pd = lambda: left.merge(right, on=\"key\", how=\"outer\")\n",
    "    pd_left_t = time_lambda(f_left_pd)\n",
    "    pd_inner_t = time_lambda(f_inner_pd)\n",
    "    pd_outer_t = time_lambda(f_outer_pd)\n",
    "    \n",
    "    # create polars dfs\n",
    "    left = pl.DataFrame(left.to_dict(orient=\"list\"))\n",
    "    right = pl.DataFrame(right.to_dict(orient=\"list\"))\n",
    "    f_left_pl = lambda: left.join(right, left_on=\"key\", right_on=\"key\", how=\"left\")\n",
    "    f_inner_pl = lambda: left.join(right, left_on=\"key\", right_on=\"key\", how=\"inner\")\n",
    "    f_outer_pl = lambda: left.join(right, left_on=\"key\", right_on=\"key\", how=\"outer\")\n",
    "    pl_left_t = time_lambda(f_left_pl)\n",
    "    pl_inner_t = time_lambda(f_inner_pl)\n",
    "    pl_outer_t = time_lambda(f_outer_pl)\n",
    "    \n",
    "    f_left_par = lambda: left.join(right, left_on=\"key\", right_on=\"key\", how=\"left\", parallel=True)\n",
    "    f_inner_par = lambda: left.join(right, left_on=\"key\", right_on=\"key\", how=\"inner\", parallel=True)\n",
    "    par_left_t = time_lambda(f_left_par)\n",
    "    par_inner_t = time_lambda(f_inner_par)\n",
    "    \n",
    "    # pandas\n",
    "    left_pd.append(pd_left_t)\n",
    "    inner_pd.append(pd_inner_t)\n",
    "    outer_pd.append(pd_outer_t)\n",
    "    \n",
    "    # polars\n",
    "    left_pl.append(pl_left_t)\n",
    "    inner_pl.append(pl_inner_t)\n",
    "    outer_pl.append(pl_outer_t)\n",
    "    \n",
    "    # parallel polars\n",
    "    par_left.append(par_left_t)\n",
    "    par_inner.append(par_inner_t)\n",
    "    n_proxy.append(N * 0.8)\n",
    "    del left\n",
    "    del right\n",
    "    \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [],
   "source": [
    "def make_fig(how):\n",
    "    plt.figure(figsize=(18, 6))\n",
    "    if how == \"inner\":\n",
    "        plt.plot(n_proxy, left_pd, label=\"pandas left\")\n",
    "        plt.plot(n_proxy, left_pl, label=\"polars left\")\n",
    "        plt.plot(n_proxy, par_left, label=\"polars left parallel\")\n",
    "    elif how == \"outer\":\n",
    "        plt.plot(n_proxy, outer_pl, label=\"polars outer\")\n",
    "        plt.plot(n_proxy, outer_pd, label=\"pandas outer\")\n",
    "    else:\n",
    "        plt.plot(n_proxy, inner_pl, label=\"polars inner\")\n",
    "        plt.plot(n_proxy, inner_pd, label=\"pandas inner\")\n",
    "        plt.plot(n_proxy, par_inner, label=\"polars inner parallel\")\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1296x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1296x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1296x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "for how in [\"inner\", \"left\", \"outer\"]:\n",
    "    make_fig(how)\n",
    "    plt.legend()\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {},
   "outputs": [],
   "source": [
    "speedup_left = np.array(left_pd) / np.array(left_pl)\n",
    "speedup_inner = np.array(inner_pd) / np.array(inner_pl)\n",
    "speedup_outer = np.array(outer_pd) / np.array(outer_pl)\n",
    "speedup_left_par = np.array(left_pd) / np.array(par_left)\n",
    "speedup_inner_par = np.array(inner_pd) / np.array(par_inner)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7effad78ca58>"
      ]
     },
     "execution_count": 96,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1224x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df = pd.DataFrame({\n",
    "    \"n_rows\": n_proxy,\n",
    "    \"speedup_left\": speedup_left,\n",
    "    \"speedup_inner\": speedup_inner,\n",
    "    \"speedup_outer\": speedup_outer,\n",
    "    \"speedup_left_par\": speedup_left,\n",
    "    \"speedup_inner_par\": speedup_inner,\n",
    "})\n",
    "df = df.set_index(\"n_rows\") \n",
    "df.plot.bar(figsize=(17, 8), ylim=(1, 3))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {},
   "outputs": [],
   "source": [
    "def create_gb_df(size):\n",
    "    str_groups = np.array(list(\"0123456789\"))\n",
    "    groups = np.arange(10)\n",
    "    g = np.random.choice(groups, size)\n",
    "    sg = np.random.choice(str_groups, size)\n",
    "    v = np.random.randn(size)\n",
    "    return pd.DataFrame({\"groups\": g, \"values\": v, \"str\": sg})\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 115,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "  0%|          | 0/3 [00:00<?, ?it/s]\u001b[A\u001b[A\n",
      "\n",
      " 33%|███▎      | 1/3 [00:00<00:00,  3.60it/s]\u001b[A\u001b[A\n",
      "\n",
      " 67%|██████▋   | 2/3 [00:01<00:00,  1.91it/s]\u001b[A\u001b[A\n",
      "\n",
      "100%|██████████| 3/3 [00:12<00:00,  3.69s/it]\u001b[A\u001b[A"
     ]
    }
   ],
   "source": [
    "gb_pd = []\n",
    "gb_pl = []\n",
    "x = []\n",
    "\n",
    "for size in tqdm.tqdm((1e2, 1e3, 1e4)):\n",
    "    size = int(1000 * size)\n",
    "    x.append(size)\n",
    "    \n",
    "    df = create_gb_df(size)\n",
    "    t = time_lambda(lambda: df.groupby(\"groups\").agg(\"str\").count())\n",
    "\n",
    "    gb_pd.append(t)\n",
    "    df = pl.DataFrame(df.to_dict(orient=\"list\"))\n",
    "    t = time_lambda(lambda: df.groupby(\"groups\").select(\"str\").count())\n",
    "    gb_pl.append(t)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 116,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7efe80782710>"
      ]
     },
     "execution_count": 116,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1224x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df = pd.DataFrame({\n",
    "    \"dataset_size\": x,\n",
    "    \"speedup\": np.array(gb_pd) / np.array(gb_pl),\n",
    "})\n",
    "df = df.set_index(\"dataset_size\")\n",
    "df.plot.bar(figsize=(17, 8))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}