toc_trends_oct_2018_part2.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "import nivapy3 as nivapy\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import os\n",
    "import seaborn as sn\n",
    "import matplotlib.pyplot as plt\n",
    "import toc_trends_analysis as resa2_trends\n",
    "import warnings\n",
    "\n",
    "warnings.filterwarnings(\"ignore\", message=\"Mean of empty slice\")\n",
    "plt.style.use('ggplot')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Connection successful.\n"
     ]
    }
   ],
   "source": [
    "# Connect to NIVABASE\n",
    "eng = nivapy.da.connect()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# TOC trends - October 2018 (Part 2: Chemistry trend analysis)\n",
    "\n",
    "The previous notebook created a new dataset for the trends analysis spanning ther period from 1990 to 2016. This notebook takes the new dataset and applies the same trends workflow as previously (from 2016 - see [here](https://nbviewer.jupyter.org/github/JamesSample/icpw/blob/master/toc_trends_oct_2016_part2.ipynb) for details)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# User input\n",
    "# Specify projects of interest\n",
    "proj_list = ['ICPW_TOCTRENDS_2018',]\n",
    "\n",
    "# Specify results folder\n",
    "res_fold = (r'../../update_autumn_2018/results')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 1. 1990 to 2012\n",
    "\n",
    "The code below is just for testing purposes: it runs the analysis from 1990 to 2012, which should be directly comparable to the data used during the 2016 analysis. The output generated below can therefore be compared to \n",
    "\n",
    "    C:\\Data\\James_Work\\Staff\\Heleen_d_W\\ICP_Waters\\TOC_Trends_Analysis_2015\\Results\\res_1990-2012.csv\n",
    "    \n",
    "and any differences *should* be due to changes in the data (which are hopefully due to finding and fixing errors)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Extracting data from RESA2...\n",
      "    Done.\n",
      "\n",
      "Converting units and applying sea-salt correction...\n",
      "    Done.\n",
      "\n",
      "Calculating statistics...\n",
      "    Data series for Al at site 38089 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38090 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38091 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38093 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for TOC at site 38101 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for EH at site 38101 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for ESO4 at site 38101 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for ECl at site 38101 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for ENO3 at site 38101 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for ESO4X at site 38101 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for ESO4_ECl at site 38101 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for ECa_EMg at site 38101 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for ECaX_EMgX at site 38101 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for ANC at site 38101 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for TOC at site 38208 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38210 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38211 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38221 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38234 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38236 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38237 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38238 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38239 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38250 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38258 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38272 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38280 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38291 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38301 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38307 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for EH at site 38494 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for EH at site 38495 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for EH at site 38496 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for EH at site 38497 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for EH at site 38498 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for EH at site 38499 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for EH at site 38500 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for EH at site 38501 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Done.\n",
      "\n",
      "Finished.\n"
     ]
    }
   ],
   "source": [
    "# Specify period of interest\n",
    "st_yr, end_yr = 1990, 2012\n",
    "\n",
    "# Build output paths\n",
    "plot_fold = os.path.join(res_fold, 'trends_plots_%s-%s' % (st_yr, end_yr))\n",
    "res_csv = os.path.join(res_fold, 'res_%s-%s.csv' % (st_yr, end_yr))\n",
    "dup_csv = os.path.join(res_fold, 'dup_%s-%s.csv' % (st_yr, end_yr))\n",
    "nd_csv = os.path.join(res_fold, 'nd_%s-%s.csv' % (st_yr, end_yr))\n",
    "\n",
    "# Run analysis \n",
    "res_df, dup_df, nd_df = resa2_trends.run_trend_analysis(proj_list, \n",
    "                                                        eng,\n",
    "                                                        st_yr=st_yr, \n",
    "                                                        end_yr=end_yr,\n",
    "                                                        plot=False, \n",
    "                                                        fold=None)\n",
    "\n",
    "# Delete mk_std_dev col as not relevant here\n",
    "del res_df['mk_std_dev']\n",
    "\n",
    "# Write output\n",
    "res_df.to_csv(res_csv, index=False)\n",
    "dup_df.to_csv(dup_csv, index=False)\n",
    "if nd_df:\n",
    "    nd_df.to_csv(nd_csv, index=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2. 1990 to 2016"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Extracting data from RESA2...\n",
      "    Done.\n",
      "\n",
      "Converting units and applying sea-salt correction...\n",
      "    Done.\n",
      "\n",
      "Calculating statistics...\n",
      "    Data series for Al at site 38089 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38090 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38091 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38093 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for TOC at site 38101 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for EH at site 38101 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for ESO4 at site 38101 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for ECl at site 38101 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for ENO3 at site 38101 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for ESO4X at site 38101 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for ESO4_ECl at site 38101 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for ECa_EMg at site 38101 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for ECaX_EMgX at site 38101 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for ANC at site 38101 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38211 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38236 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38237 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38238 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38239 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38250 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38258 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38272 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38280 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38291 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38301 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38307 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for EH at site 38494 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for EH at site 38495 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for EH at site 38496 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for EH at site 38497 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for EH at site 38498 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for EH at site 38499 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for EH at site 38500 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for EH at site 38501 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Done.\n",
      "\n",
      "Finished.\n"
     ]
    }
   ],
   "source": [
    "# Specify period of interest\n",
    "st_yr, end_yr = 1990, 2016\n",
    "\n",
    "# Build output paths\n",
    "plot_fold = os.path.join(res_fold, 'trends_plots_%s-%s' % (st_yr, end_yr))\n",
    "res_csv = os.path.join(res_fold, 'res_%s-%s.csv' % (st_yr, end_yr))\n",
    "dup_csv = os.path.join(res_fold, 'dup_%s-%s.csv' % (st_yr, end_yr))\n",
    "nd_csv = os.path.join(res_fold, 'nd_%s-%s.csv' % (st_yr, end_yr))\n",
    "\n",
    "# Run analysis \n",
    "res_df, dup_df, nd_df = resa2_trends.run_trend_analysis(proj_list, \n",
    "                                                        eng,\n",
    "                                                        st_yr=st_yr, \n",
    "                                                        end_yr=end_yr,\n",
    "                                                        plot=False, \n",
    "                                                        fold=None)\n",
    "\n",
    "# Delete mk_std_dev col as not relevant here\n",
    "del res_df['mk_std_dev']\n",
    "\n",
    "# Write output\n",
    "res_df.to_csv(res_csv, index=False)\n",
    "dup_df.to_csv(dup_csv, index=False)\n",
    "if nd_df:\n",
    "    nd_df.to_csv(nd_csv, index=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3. 2002 to 2016"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Extracting data from RESA2...\n",
      "    Done.\n",
      "\n",
      "Converting units and applying sea-salt correction...\n",
      "    Done.\n",
      "\n",
      "Calculating statistics...\n",
      "    Data series for Al at site 38089 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38090 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38091 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38092 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38093 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38095 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38096 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38097 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38098 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for TOC at site 38101 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for EH at site 38101 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for ESO4 at site 38101 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for ECl at site 38101 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for ENO3 at site 38101 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for ESO4X at site 38101 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for ESO4_ECl at site 38101 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38210 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38211 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38218 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38221 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38222 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38227 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38231 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for TOC at site 38231 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for EH at site 38231 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for ESO4 at site 38231 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for ECl at site 38231 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for ENO3 at site 38231 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for ESO4X at site 38231 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for ESO4_ECl at site 38231 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for ECa_EMg at site 38231 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for ECaX_EMgX at site 38231 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for ANC at site 38231 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38232 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38233 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38234 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38236 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38237 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38238 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38239 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38250 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38258 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38272 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38280 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38291 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38301 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38307 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for EH at site 38494 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for EH at site 38495 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for EH at site 38496 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for EH at site 38497 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for EH at site 38498 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for EH at site 38499 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for EH at site 38500 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for EH at site 38501 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for TOC at site 38607 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Done.\n",
      "\n",
      "Finished.\n"
     ]
    }
   ],
   "source": [
    "# Specify period of interest\n",
    "st_yr, end_yr = 2002, 2016\n",
    "\n",
    "# Build output paths\n",
    "plot_fold = os.path.join(res_fold, 'trends_plots_%s-%s' % (st_yr, end_yr))\n",
    "res_csv = os.path.join(res_fold, 'res_%s-%s.csv' % (st_yr, end_yr))\n",
    "dup_csv = os.path.join(res_fold, 'dup_%s-%s.csv' % (st_yr, end_yr))\n",
    "nd_csv = os.path.join(res_fold, 'nd_%s-%s.csv' % (st_yr, end_yr))\n",
    "\n",
    "# Run analysis \n",
    "res_df, dup_df, nd_df = resa2_trends.run_trend_analysis(proj_list, \n",
    "                                                        eng,\n",
    "                                                        st_yr=st_yr, \n",
    "                                                        end_yr=end_yr,\n",
    "                                                        plot=False, \n",
    "                                                        fold=None)\n",
    "\n",
    "# Delete mk_std_dev col as not relevant here\n",
    "del res_df['mk_std_dev']\n",
    "\n",
    "# Write output\n",
    "res_df.to_csv(res_csv, index=False)\n",
    "dup_df.to_csv(dup_csv, index=False)\n",
    "if nd_df:\n",
    "    nd_df.to_csv(nd_csv, index=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 4. 1990 to 2004"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Extracting data from RESA2...\n",
      "    Done.\n",
      "\n",
      "Converting units and applying sea-salt correction...\n",
      "    Done.\n",
      "\n",
      "Calculating statistics...\n",
      "    Data series for Al at site 38089 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38090 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38091 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38092 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38093 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for TOC at site 38101 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for EH at site 38101 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for ESO4 at site 38101 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for ECl at site 38101 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for ENO3 at site 38101 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for ESO4X at site 38101 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for ESO4_ECl at site 38101 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for ECa_EMg at site 38101 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for ECaX_EMgX at site 38101 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for ANC at site 38101 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38209 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38210 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38211 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38212 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38213 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38216 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38221 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38224 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38225 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38229 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38230 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38234 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for ENO3 at site 38426 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for ANC at site 38426 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for TOC at site 38432 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for EH at site 38432 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for ESO4 at site 38432 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for ECl at site 38432 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for ENO3 at site 38432 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for ESO4X at site 38432 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for ESO4_ECl at site 38432 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for ECa_EMg at site 38432 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for ECaX_EMgX at site 38432 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for ANC at site 38432 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for TOC at site 38498 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Done.\n",
      "\n",
      "Finished.\n"
     ]
    }
   ],
   "source": [
    "# Specify period of interest\n",
    "st_yr, end_yr = 1990, 2004\n",
    "\n",
    "# Build output paths\n",
    "plot_fold = os.path.join(res_fold, 'trends_plots_%s-%s' % (st_yr, end_yr))\n",
    "res_csv = os.path.join(res_fold, 'res_%s-%s.csv' % (st_yr, end_yr))\n",
    "dup_csv = os.path.join(res_fold, 'dup_%s-%s.csv' % (st_yr, end_yr))\n",
    "nd_csv = os.path.join(res_fold, 'nd_%s-%s.csv' % (st_yr, end_yr))\n",
    "\n",
    "# Run analysis \n",
    "res_df, dup_df, nd_df = resa2_trends.run_trend_analysis(proj_list, \n",
    "                                                        eng,\n",
    "                                                        st_yr=st_yr, \n",
    "                                                        end_yr=end_yr,\n",
    "                                                        plot=False, \n",
    "                                                        fold=None)\n",
    "\n",
    "# Delete mk_std_dev col as not relevant here\n",
    "del res_df['mk_std_dev']\n",
    "\n",
    "# Write output\n",
    "res_df.to_csv(res_csv, index=False)\n",
    "dup_df.to_csv(dup_csv, index=False)\n",
    "if nd_df:\n",
    "    nd_df.to_csv(nd_csv, index=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 5. All data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Extracting data from RESA2...\n",
      "    Done.\n",
      "\n",
      "Converting units and applying sea-salt correction...\n",
      "    Done.\n",
      "\n",
      "Calculating statistics...\n",
      "    Data series for TOC at site 38101 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for EH at site 38101 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for ESO4 at site 38101 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for ECl at site 38101 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for ENO3 at site 38101 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for ESO4X at site 38101 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for ESO4_ECl at site 38101 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for ECa_EMg at site 38101 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for ECaX_EMgX at site 38101 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for ANC at site 38101 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38211 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38236 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38237 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38238 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38239 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38250 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38258 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38272 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38280 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38291 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38301 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for Al at site 38307 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for EH at site 38494 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for EH at site 38495 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for EH at site 38496 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for EH at site 38497 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for EH at site 38498 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for EH at site 38499 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for EH at site 38500 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Data series for EH at site 38501 has fewer than 10 non-null values. Significance estimates may be unreliable.\n",
      "    Done.\n",
      "\n",
      "Finished.\n"
     ]
    }
   ],
   "source": [
    "# Specify period of interest\n",
    "st_yr, end_yr = None, None\n",
    "\n",
    "# Build output paths\n",
    "plot_fold = os.path.join(res_fold, 'trends_plots_all_years')\n",
    "res_csv = os.path.join(res_fold, 'res_all_years.csv')\n",
    "dup_csv = os.path.join(res_fold, 'dup_all_years.csv')\n",
    "nd_csv = os.path.join(res_fold, 'nd_all_years.csv')\n",
    "\n",
    "# Run analysis \n",
    "res_df, dup_df, nd_df = resa2_trends.run_trend_analysis(proj_list, \n",
    "                                                        eng,\n",
    "                                                        st_yr=st_yr, \n",
    "                                                        end_yr=end_yr,\n",
    "                                                        plot=False, \n",
    "                                                        fold=None)\n",
    "\n",
    "# Delete mk_std_dev col as not relevant here\n",
    "del res_df['mk_std_dev']\n",
    "\n",
    "# Write output\n",
    "res_df.to_csv(res_csv, index=False)\n",
    "dup_df.to_csv(dup_csv, index=False)\n",
    "if nd_df:\n",
    "    nd_df.to_csv(nd_csv, index=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 6. Basic checking\n",
    "\n",
    "### 6.1. Boxplots\n",
    "\n",
    "The plot below can be compared to the previous results [here](https://nbviewer.jupyter.org/github/JamesSample/icpw/blob/master/toc_trends_oct_2016_part2.ipynb#6.1.-Boxplots). Overall, the two plots are very similar, which is reassuring."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Set up plot\n",
    "fig = plt.figure(figsize=(20,10))\n",
    "sn.set(style=\"ticks\", palette=\"muted\", \n",
    "       color_codes=True, font_scale=2)\n",
    "\n",
    "# Horizontal boxplots\n",
    "ax = sn.boxplot(x=\"mean\", y=\"par_id\", data=res_df,\n",
    "                whis=np.inf, color=\"c\")\n",
    "\n",
    "# Add \"raw\" data points for each observation, with some \"jitter\"\n",
    "# to make them visible\n",
    "sn.stripplot(x=\"mean\", y=\"par_id\", data=res_df, jitter=True, \n",
    "             size=3, color=\".3\", linewidth=0)\n",
    "\n",
    "# Remove axis lines\n",
    "sn.despine(trim=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 7. Data restructuring\n",
    "\n",
    "The code below is taken from [here](https://nbviewer.jupyter.org/github/JamesSample/icpw/blob/master/toc_trends_oct_2016_part2.ipynb#7.-Data-restructuring). It is used to generate output files in the format requested by Heleen.\n",
    "\n",
    "### 7.1. Combine datasets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "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>project_id</th>\n",
       "      <th>project_name</th>\n",
       "      <th>station_id</th>\n",
       "      <th>station_code</th>\n",
       "      <th>station_name</th>\n",
       "      <th>nfc_code</th>\n",
       "      <th>type</th>\n",
       "      <th>latitude</th>\n",
       "      <th>longitude</th>\n",
       "      <th>continent</th>\n",
       "      <th>...</th>\n",
       "      <th>n_start</th>\n",
       "      <th>n_end</th>\n",
       "      <th>mean</th>\n",
       "      <th>median</th>\n",
       "      <th>std_dev</th>\n",
       "      <th>mk_stat</th>\n",
       "      <th>norm_mk_stat</th>\n",
       "      <th>mk_p_val</th>\n",
       "      <th>trend</th>\n",
       "      <th>sen_slp</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>4390</td>\n",
       "      <td>ICPW_TOCTRENDS_2018</td>\n",
       "      <td>38085</td>\n",
       "      <td>Tr18_CA01</td>\n",
       "      <td>Ontario, Algoma Region, Batchawana Lake</td>\n",
       "      <td></td>\n",
       "      <td>L</td>\n",
       "      <td>47.06</td>\n",
       "      <td>-84.393</td>\n",
       "      <td>NA</td>\n",
       "      <td>...</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>102.129696</td>\n",
       "      <td>104.000000</td>\n",
       "      <td>19.108938</td>\n",
       "      <td>19.0</td>\n",
       "      <td>0.375244</td>\n",
       "      <td>7.074787e-01</td>\n",
       "      <td>no trend</td>\n",
       "      <td>0.121250</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>4390</td>\n",
       "      <td>ICPW_TOCTRENDS_2018</td>\n",
       "      <td>38085</td>\n",
       "      <td>Tr18_CA01</td>\n",
       "      <td>Ontario, Algoma Region, Batchawana Lake</td>\n",
       "      <td></td>\n",
       "      <td>L</td>\n",
       "      <td>47.06</td>\n",
       "      <td>-84.393</td>\n",
       "      <td>NA</td>\n",
       "      <td>...</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>4.703870</td>\n",
       "      <td>4.668000</td>\n",
       "      <td>0.319514</td>\n",
       "      <td>142.0</td>\n",
       "      <td>2.940053</td>\n",
       "      <td>3.281559e-03</td>\n",
       "      <td>increasing</td>\n",
       "      <td>0.019687</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>4390</td>\n",
       "      <td>ICPW_TOCTRENDS_2018</td>\n",
       "      <td>38085</td>\n",
       "      <td>Tr18_CA01</td>\n",
       "      <td>Ontario, Algoma Region, Batchawana Lake</td>\n",
       "      <td></td>\n",
       "      <td>L</td>\n",
       "      <td>47.06</td>\n",
       "      <td>-84.393</td>\n",
       "      <td>NA</td>\n",
       "      <td>...</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>1.049825</td>\n",
       "      <td>1.023293</td>\n",
       "      <td>0.251578</td>\n",
       "      <td>-38.0</td>\n",
       "      <td>-0.772287</td>\n",
       "      <td>4.399443e-01</td>\n",
       "      <td>no trend</td>\n",
       "      <td>-0.005823</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4390</td>\n",
       "      <td>ICPW_TOCTRENDS_2018</td>\n",
       "      <td>38085</td>\n",
       "      <td>Tr18_CA01</td>\n",
       "      <td>Ontario, Algoma Region, Batchawana Lake</td>\n",
       "      <td></td>\n",
       "      <td>L</td>\n",
       "      <td>47.06</td>\n",
       "      <td>-84.393</td>\n",
       "      <td>NA</td>\n",
       "      <td>...</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>84.743245</td>\n",
       "      <td>95.000000</td>\n",
       "      <td>20.025680</td>\n",
       "      <td>-289.0</td>\n",
       "      <td>-6.003910</td>\n",
       "      <td>1.926214e-09</td>\n",
       "      <td>decreasing</td>\n",
       "      <td>-2.333912</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4390</td>\n",
       "      <td>ICPW_TOCTRENDS_2018</td>\n",
       "      <td>38085</td>\n",
       "      <td>Tr18_CA01</td>\n",
       "      <td>Ontario, Algoma Region, Batchawana Lake</td>\n",
       "      <td></td>\n",
       "      <td>L</td>\n",
       "      <td>47.06</td>\n",
       "      <td>-84.393</td>\n",
       "      <td>NA</td>\n",
       "      <td>...</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>7.166138</td>\n",
       "      <td>7.142857</td>\n",
       "      <td>1.951420</td>\n",
       "      <td>-135.0</td>\n",
       "      <td>-2.811464</td>\n",
       "      <td>4.931664e-03</td>\n",
       "      <td>decreasing</td>\n",
       "      <td>-0.155429</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 27 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   project_id         project_name  station_id station_code  \\\n",
       "0        4390  ICPW_TOCTRENDS_2018       38085    Tr18_CA01   \n",
       "1        4390  ICPW_TOCTRENDS_2018       38085    Tr18_CA01   \n",
       "2        4390  ICPW_TOCTRENDS_2018       38085    Tr18_CA01   \n",
       "3        4390  ICPW_TOCTRENDS_2018       38085    Tr18_CA01   \n",
       "4        4390  ICPW_TOCTRENDS_2018       38085    Tr18_CA01   \n",
       "\n",
       "                              station_name nfc_code type  latitude  longitude  \\\n",
       "0  Ontario, Algoma Region, Batchawana Lake             L     47.06    -84.393   \n",
       "1  Ontario, Algoma Region, Batchawana Lake             L     47.06    -84.393   \n",
       "2  Ontario, Algoma Region, Batchawana Lake             L     47.06    -84.393   \n",
       "3  Ontario, Algoma Region, Batchawana Lake             L     47.06    -84.393   \n",
       "4  Ontario, Algoma Region, Batchawana Lake             L     47.06    -84.393   \n",
       "\n",
       "  continent    ...    n_start n_end        mean      median    std_dev  \\\n",
       "0        NA    ...          5     5  102.129696  104.000000  19.108938   \n",
       "1        NA    ...          5     5    4.703870    4.668000   0.319514   \n",
       "2        NA    ...          5     5    1.049825    1.023293   0.251578   \n",
       "3        NA    ...          5     5   84.743245   95.000000  20.025680   \n",
       "4        NA    ...          5     5    7.166138    7.142857   1.951420   \n",
       "\n",
       "  mk_stat  norm_mk_stat      mk_p_val       trend   sen_slp  \n",
       "0    19.0      0.375244  7.074787e-01    no trend  0.121250  \n",
       "1   142.0      2.940053  3.281559e-03  increasing  0.019687  \n",
       "2   -38.0     -0.772287  4.399443e-01    no trend -0.005823  \n",
       "3  -289.0     -6.003910  1.926214e-09  decreasing -2.333912  \n",
       "4  -135.0     -2.811464  4.931664e-03  decreasing -0.155429  \n",
       "\n",
       "[5 rows x 27 columns]"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Read results files and concatenate\n",
    "# Container for data\n",
    "df_list = []\n",
    "\n",
    "# Loop over periods\n",
    "for per in ['1990-2016', '1990-2004', '2002-2016', 'all_years']:\n",
    "    res_path = os.path.join(res_fold, 'res_%s.csv' % per)\n",
    "    df = pd.read_csv(res_path)\n",
    "    \n",
    "    # Change 'period' col to 'data_period' and add 'analysis_period'\n",
    "    df['data_period'] = df['period']\n",
    "    del df['period']\n",
    "    \n",
    "    df['analysis_period'] = per\n",
    "    \n",
    "    df_list.append(df)\n",
    "    \n",
    "# Concat\n",
    "df = pd.concat(df_list, axis=0)\n",
    "\n",
    "# Read station data\n",
    "stn_path = r'../../update_autumn_2018/toc_trends_oct18_stations.xlsx'\n",
    "stn_df = pd.read_excel(stn_path, sheet_name='Data', keep_default_na=False)\n",
    "\n",
    "# Join\n",
    "df = pd.merge(df, stn_df, how='left', on='station_id')\n",
    "\n",
    "# Read projects table\n",
    "sql = (\"SELECT project_id, project_name \"\n",
    "       \"FROM resa2.projects \"\n",
    "       \"WHERE project_name = 'ICPW_TOCTRENDS_2018'\")\n",
    "proj_df = pd.read_sql_query(sql, eng)\n",
    "\n",
    "# Get associated stations\n",
    "sql = (\"SELECT station_id, project_id \"\n",
    "       \"FROM resa2.projects_stations \"\n",
    "       \"WHERE project_id = 4390\")\n",
    "proj_stn_df = pd.read_sql_query(sql, eng)\n",
    "\n",
    "# Join proj details\n",
    "proj_df = pd.merge(proj_stn_df, proj_df, how='left', on ='project_id')\n",
    "\n",
    "# Join to results\n",
    "df = pd.merge(df, proj_df, how='left', on='station_id')\n",
    "\n",
    "# Re-order columns\n",
    "df = df[['project_id', 'project_name', 'station_id', \n",
    "         'station_code', 'station_name', 'nfc_code', 'type',\n",
    "         'latitude', 'longitude', 'continent', 'country', \n",
    "         'region', 'subregion', 'analysis_period', 'data_period', \n",
    "         'par_id', 'non_missing', 'n_start', 'n_end', 'mean', 'median', \n",
    "         'std_dev', 'mk_stat', 'norm_mk_stat', 'mk_p_val', 'trend',\n",
    "         'sen_slp']]\n",
    "\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 7.2. Check record completeness\n",
    "\n",
    "See e-mail from Heleen received 25/10/2016 at 15:56. The `'non_missing'` threshold is based of 65% of the data period (e.g. 65% of 27 years for 1990 to 2016)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "def include(row):\n",
    "    if ((row['analysis_period'] == '1990-2016') & \n",
    "        (row['n_start'] >= 2) &\n",
    "        (row['n_end'] >= 2) &\n",
    "        (row['non_missing'] >= 18)):\n",
    "        return 'yes'\n",
    "    elif ((row['analysis_period'] == '1990-2004') & \n",
    "          (row['n_start'] >= 2) &\n",
    "          (row['n_end'] >= 2) &          \n",
    "          (row['non_missing'] >= 10)):\n",
    "        return 'yes'\n",
    "    elif ((row['analysis_period'] == '2002-2016') & \n",
    "          (row['n_start'] >= 2) &\n",
    "          (row['n_end'] >= 2) &          \n",
    "          (row['non_missing'] >= 10)):\n",
    "        return 'yes'\n",
    "    else:\n",
    "        return 'no'\n",
    "\n",
    "df['include'] = df.apply(include, axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 7.3. SO4 at Abiskojaure\n",
    "\n",
    "SO4 for this station (`'station_id=38335'`) should be removed. See [here](https://nbviewer.jupyter.org/github/JamesSample/icpw/blob/master/toc_trends_oct_2016_part3.ipynb#1.3.-SO4-for-station-ID-36561)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Remove sulphate-related series at Abiskojaure\n",
    "df = df.query('not((station_id==38335) and ((par_id==\"ESO4\") or '\n",
    "                                           '(par_id==\"ESO4X\") or '\n",
    "                                           '(par_id==\"ESO4_ECl\")))')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 7.4. Relative slope"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Relative slope\n",
    "df['rel_sen_slp'] = df['sen_slp'] / df['median'] "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 7.5. Tidy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "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>project_id</th>\n",
       "      <th>project_name</th>\n",
       "      <th>station_id</th>\n",
       "      <th>station_code</th>\n",
       "      <th>station_name</th>\n",
       "      <th>nfc_code</th>\n",
       "      <th>type</th>\n",
       "      <th>continent</th>\n",
       "      <th>country</th>\n",
       "      <th>region</th>\n",
       "      <th>...</th>\n",
       "      <th>data_period</th>\n",
       "      <th>par_id</th>\n",
       "      <th>non_missing</th>\n",
       "      <th>median</th>\n",
       "      <th>std_dev</th>\n",
       "      <th>mk_p_val</th>\n",
       "      <th>trend</th>\n",
       "      <th>sen_slp</th>\n",
       "      <th>rel_sen_slp</th>\n",
       "      <th>include</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>4390</td>\n",
       "      <td>ICPW_TOCTRENDS_2018</td>\n",
       "      <td>38085</td>\n",
       "      <td>Tr18_CA01</td>\n",
       "      <td>Ontario, Algoma Region, Batchawana Lake</td>\n",
       "      <td></td>\n",
       "      <td>L</td>\n",
       "      <td>NA</td>\n",
       "      <td>Canada</td>\n",
       "      <td>ONT</td>\n",
       "      <td>...</td>\n",
       "      <td>1990-2016</td>\n",
       "      <td>Al</td>\n",
       "      <td>27</td>\n",
       "      <td>104.000000</td>\n",
       "      <td>19.108938</td>\n",
       "      <td>7.074787e-01</td>\n",
       "      <td>no trend</td>\n",
       "      <td>0.121250</td>\n",
       "      <td>0.001166</td>\n",
       "      <td>yes</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>4390</td>\n",
       "      <td>ICPW_TOCTRENDS_2018</td>\n",
       "      <td>38085</td>\n",
       "      <td>Tr18_CA01</td>\n",
       "      <td>Ontario, Algoma Region, Batchawana Lake</td>\n",
       "      <td></td>\n",
       "      <td>L</td>\n",
       "      <td>NA</td>\n",
       "      <td>Canada</td>\n",
       "      <td>ONT</td>\n",
       "      <td>...</td>\n",
       "      <td>1990-2016</td>\n",
       "      <td>TOC</td>\n",
       "      <td>27</td>\n",
       "      <td>4.668000</td>\n",
       "      <td>0.319514</td>\n",
       "      <td>3.281559e-03</td>\n",
       "      <td>increasing</td>\n",
       "      <td>0.019687</td>\n",
       "      <td>0.004218</td>\n",
       "      <td>yes</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>4390</td>\n",
       "      <td>ICPW_TOCTRENDS_2018</td>\n",
       "      <td>38085</td>\n",
       "      <td>Tr18_CA01</td>\n",
       "      <td>Ontario, Algoma Region, Batchawana Lake</td>\n",
       "      <td></td>\n",
       "      <td>L</td>\n",
       "      <td>NA</td>\n",
       "      <td>Canada</td>\n",
       "      <td>ONT</td>\n",
       "      <td>...</td>\n",
       "      <td>1990-2016</td>\n",
       "      <td>EH</td>\n",
       "      <td>27</td>\n",
       "      <td>1.023293</td>\n",
       "      <td>0.251578</td>\n",
       "      <td>4.399443e-01</td>\n",
       "      <td>no trend</td>\n",
       "      <td>-0.005823</td>\n",
       "      <td>-0.005691</td>\n",
       "      <td>yes</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4390</td>\n",
       "      <td>ICPW_TOCTRENDS_2018</td>\n",
       "      <td>38085</td>\n",
       "      <td>Tr18_CA01</td>\n",
       "      <td>Ontario, Algoma Region, Batchawana Lake</td>\n",
       "      <td></td>\n",
       "      <td>L</td>\n",
       "      <td>NA</td>\n",
       "      <td>Canada</td>\n",
       "      <td>ONT</td>\n",
       "      <td>...</td>\n",
       "      <td>1990-2016</td>\n",
       "      <td>ESO4</td>\n",
       "      <td>27</td>\n",
       "      <td>95.000000</td>\n",
       "      <td>20.025680</td>\n",
       "      <td>1.926214e-09</td>\n",
       "      <td>decreasing</td>\n",
       "      <td>-2.333912</td>\n",
       "      <td>-0.024567</td>\n",
       "      <td>yes</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4390</td>\n",
       "      <td>ICPW_TOCTRENDS_2018</td>\n",
       "      <td>38085</td>\n",
       "      <td>Tr18_CA01</td>\n",
       "      <td>Ontario, Algoma Region, Batchawana Lake</td>\n",
       "      <td></td>\n",
       "      <td>L</td>\n",
       "      <td>NA</td>\n",
       "      <td>Canada</td>\n",
       "      <td>ONT</td>\n",
       "      <td>...</td>\n",
       "      <td>1990-2016</td>\n",
       "      <td>ECl</td>\n",
       "      <td>27</td>\n",
       "      <td>7.142857</td>\n",
       "      <td>1.951420</td>\n",
       "      <td>4.931664e-03</td>\n",
       "      <td>decreasing</td>\n",
       "      <td>-0.155429</td>\n",
       "      <td>-0.021760</td>\n",
       "      <td>yes</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 24 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   project_id         project_name  station_id station_code  \\\n",
       "0        4390  ICPW_TOCTRENDS_2018       38085    Tr18_CA01   \n",
       "1        4390  ICPW_TOCTRENDS_2018       38085    Tr18_CA01   \n",
       "2        4390  ICPW_TOCTRENDS_2018       38085    Tr18_CA01   \n",
       "3        4390  ICPW_TOCTRENDS_2018       38085    Tr18_CA01   \n",
       "4        4390  ICPW_TOCTRENDS_2018       38085    Tr18_CA01   \n",
       "\n",
       "                              station_name nfc_code type continent country  \\\n",
       "0  Ontario, Algoma Region, Batchawana Lake             L        NA  Canada   \n",
       "1  Ontario, Algoma Region, Batchawana Lake             L        NA  Canada   \n",
       "2  Ontario, Algoma Region, Batchawana Lake             L        NA  Canada   \n",
       "3  Ontario, Algoma Region, Batchawana Lake             L        NA  Canada   \n",
       "4  Ontario, Algoma Region, Batchawana Lake             L        NA  Canada   \n",
       "\n",
       "  region   ...    data_period  par_id  non_missing      median    std_dev  \\\n",
       "0    ONT   ...      1990-2016      Al           27  104.000000  19.108938   \n",
       "1    ONT   ...      1990-2016     TOC           27    4.668000   0.319514   \n",
       "2    ONT   ...      1990-2016      EH           27    1.023293   0.251578   \n",
       "3    ONT   ...      1990-2016    ESO4           27   95.000000  20.025680   \n",
       "4    ONT   ...      1990-2016     ECl           27    7.142857   1.951420   \n",
       "\n",
       "       mk_p_val       trend   sen_slp  rel_sen_slp  include  \n",
       "0  7.074787e-01    no trend  0.121250     0.001166      yes  \n",
       "1  3.281559e-03  increasing  0.019687     0.004218      yes  \n",
       "2  4.399443e-01    no trend -0.005823    -0.005691      yes  \n",
       "3  1.926214e-09  decreasing -2.333912    -0.024567      yes  \n",
       "4  4.931664e-03  decreasing -0.155429    -0.021760      yes  \n",
       "\n",
       "[5 rows x 24 columns]"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Remove unwanted cols\n",
    "df.drop(labels=['mean', 'n_end', 'n_start', 'mk_stat', 'norm_mk_stat'],\n",
    "        axis=1, inplace=True)\n",
    "\n",
    "# Reorder columns\n",
    "df = df[['project_id', 'project_name', 'station_id', 'station_code', \n",
    "         'station_name', 'nfc_code', 'type', 'continent', 'country', \n",
    "         'region', 'subregion', 'latitude', 'longitude', 'analysis_period',\n",
    "         'data_period', 'par_id', 'non_missing', 'median', 'std_dev',\n",
    "         'mk_p_val', 'trend', 'sen_slp', 'rel_sen_slp', 'include']]\n",
    "\n",
    "# Write to output\n",
    "out_path = r'../../update_autumn_2018/results/toc_trends_long_format.csv'\n",
    "df.to_csv(out_path, index=False, encoding='utf-8')\n",
    "\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 7.6. Convert to \"wide\" format"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "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>project_id</th>\n",
       "      <th>project_name</th>\n",
       "      <th>station_id</th>\n",
       "      <th>station_code</th>\n",
       "      <th>station_name</th>\n",
       "      <th>nfc_code</th>\n",
       "      <th>type</th>\n",
       "      <th>continent</th>\n",
       "      <th>country</th>\n",
       "      <th>region</th>\n",
       "      <th>...</th>\n",
       "      <th>TOC_2002-2016_rel_sen_slp</th>\n",
       "      <th>EH_2002-2016_rel_sen_slp</th>\n",
       "      <th>ESO4_2002-2016_rel_sen_slp</th>\n",
       "      <th>ECl_2002-2016_rel_sen_slp</th>\n",
       "      <th>ENO3_2002-2016_rel_sen_slp</th>\n",
       "      <th>ESO4X_2002-2016_rel_sen_slp</th>\n",
       "      <th>ESO4_ECl_2002-2016_rel_sen_slp</th>\n",
       "      <th>ECa_EMg_2002-2016_rel_sen_slp</th>\n",
       "      <th>ECaX_EMgX_2002-2016_rel_sen_slp</th>\n",
       "      <th>ANC_2002-2016_rel_sen_slp</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>4390</td>\n",
       "      <td>ICPW_TOCTRENDS_2018</td>\n",
       "      <td>38085</td>\n",
       "      <td>Tr18_CA01</td>\n",
       "      <td>Ontario, Algoma Region, Batchawana Lake</td>\n",
       "      <td></td>\n",
       "      <td>L</td>\n",
       "      <td>NA</td>\n",
       "      <td>Canada</td>\n",
       "      <td>ONT</td>\n",
       "      <td>...</td>\n",
       "      <td>0.00370616</td>\n",
       "      <td>-0.0283587</td>\n",
       "      <td>-0.0568098</td>\n",
       "      <td>-0.0452273</td>\n",
       "      <td>-0.015261</td>\n",
       "      <td>-0.0569253</td>\n",
       "      <td>-0.0575711</td>\n",
       "      <td>-0.0102527</td>\n",
       "      <td>-0.00992399</td>\n",
       "      <td>0.0247881</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>4390</td>\n",
       "      <td>ICPW_TOCTRENDS_2018</td>\n",
       "      <td>38086</td>\n",
       "      <td>Tr18_CA02</td>\n",
       "      <td>Ontario, Algoma Region, Wishart Lake</td>\n",
       "      <td></td>\n",
       "      <td>L</td>\n",
       "      <td>NA</td>\n",
       "      <td>Canada</td>\n",
       "      <td>ONT</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000662617</td>\n",
       "      <td>-0.0170369</td>\n",
       "      <td>-0.0434934</td>\n",
       "      <td>-0.0343182</td>\n",
       "      <td>-0.00547504</td>\n",
       "      <td>-0.0432656</td>\n",
       "      <td>-0.0430517</td>\n",
       "      <td>-0.0101394</td>\n",
       "      <td>-0.0100083</td>\n",
       "      <td>0.0166807</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>4390</td>\n",
       "      <td>ICPW_TOCTRENDS_2018</td>\n",
       "      <td>38087</td>\n",
       "      <td>Tr18_CA03</td>\n",
       "      <td>Ontario, Algoma Region, Little Turkey Lake</td>\n",
       "      <td></td>\n",
       "      <td>L</td>\n",
       "      <td>NA</td>\n",
       "      <td>Canada</td>\n",
       "      <td>ONT</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000395361</td>\n",
       "      <td>-0.013686</td>\n",
       "      <td>-0.0376313</td>\n",
       "      <td>-0.045</td>\n",
       "      <td>-0.00533808</td>\n",
       "      <td>-0.0371517</td>\n",
       "      <td>-0.0389343</td>\n",
       "      <td>-0.00722479</td>\n",
       "      <td>-0.00720022</td>\n",
       "      <td>0.0106598</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4390</td>\n",
       "      <td>ICPW_TOCTRENDS_2018</td>\n",
       "      <td>38088</td>\n",
       "      <td>Tr18_CA04</td>\n",
       "      <td>Ontario, Algoma Region, Turkey Lake</td>\n",
       "      <td></td>\n",
       "      <td>L</td>\n",
       "      <td>NA</td>\n",
       "      <td>Canada</td>\n",
       "      <td>ONT</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.00139743</td>\n",
       "      <td>-0.0167356</td>\n",
       "      <td>-0.0343655</td>\n",
       "      <td>-0.043379</td>\n",
       "      <td>0.000243665</td>\n",
       "      <td>-0.0344036</td>\n",
       "      <td>-0.0367411</td>\n",
       "      <td>0.000580972</td>\n",
       "      <td>0.000841425</td>\n",
       "      <td>0.0142161</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4390</td>\n",
       "      <td>ICPW_TOCTRENDS_2018</td>\n",
       "      <td>38089</td>\n",
       "      <td>Tr18_CA05</td>\n",
       "      <td>Quebec, Lac Veilleux</td>\n",
       "      <td></td>\n",
       "      <td>L</td>\n",
       "      <td>NA</td>\n",
       "      <td>Canada</td>\n",
       "      <td>QuMaVt</td>\n",
       "      <td>...</td>\n",
       "      <td>0.00512821</td>\n",
       "      <td>-0.0171366</td>\n",
       "      <td>-0.0421053</td>\n",
       "      <td>-0.0888889</td>\n",
       "      <td>-0.111131</td>\n",
       "      <td>-0.035929</td>\n",
       "      <td>-0.0415389</td>\n",
       "      <td>-0.012619</td>\n",
       "      <td>-0.00857364</td>\n",
       "      <td>0.017423</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 244 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   project_id         project_name  station_id station_code  \\\n",
       "0        4390  ICPW_TOCTRENDS_2018       38085    Tr18_CA01   \n",
       "1        4390  ICPW_TOCTRENDS_2018       38086    Tr18_CA02   \n",
       "2        4390  ICPW_TOCTRENDS_2018       38087    Tr18_CA03   \n",
       "3        4390  ICPW_TOCTRENDS_2018       38088    Tr18_CA04   \n",
       "4        4390  ICPW_TOCTRENDS_2018       38089    Tr18_CA05   \n",
       "\n",
       "                                 station_name nfc_code type continent country  \\\n",
       "0     Ontario, Algoma Region, Batchawana Lake             L        NA  Canada   \n",
       "1        Ontario, Algoma Region, Wishart Lake             L        NA  Canada   \n",
       "2  Ontario, Algoma Region, Little Turkey Lake             L        NA  Canada   \n",
       "3         Ontario, Algoma Region, Turkey Lake             L        NA  Canada   \n",
       "4                        Quebec, Lac Veilleux             L        NA  Canada   \n",
       "\n",
       "   region            ...            TOC_2002-2016_rel_sen_slp  \\\n",
       "0     ONT            ...                           0.00370616   \n",
       "1     ONT            ...                          0.000662617   \n",
       "2     ONT            ...                          0.000395361   \n",
       "3     ONT            ...                          -0.00139743   \n",
       "4  QuMaVt            ...                           0.00512821   \n",
       "\n",
       "   EH_2002-2016_rel_sen_slp  ESO4_2002-2016_rel_sen_slp  \\\n",
       "0                -0.0283587                  -0.0568098   \n",
       "1                -0.0170369                  -0.0434934   \n",
       "2                 -0.013686                  -0.0376313   \n",
       "3                -0.0167356                  -0.0343655   \n",
       "4                -0.0171366                  -0.0421053   \n",
       "\n",
       "  ECl_2002-2016_rel_sen_slp ENO3_2002-2016_rel_sen_slp  \\\n",
       "0                -0.0452273                  -0.015261   \n",
       "1                -0.0343182                -0.00547504   \n",
       "2                    -0.045                -0.00533808   \n",
       "3                 -0.043379                0.000243665   \n",
       "4                -0.0888889                  -0.111131   \n",
       "\n",
       "  ESO4X_2002-2016_rel_sen_slp ESO4_ECl_2002-2016_rel_sen_slp  \\\n",
       "0                  -0.0569253                     -0.0575711   \n",
       "1                  -0.0432656                     -0.0430517   \n",
       "2                  -0.0371517                     -0.0389343   \n",
       "3                  -0.0344036                     -0.0367411   \n",
       "4                   -0.035929                     -0.0415389   \n",
       "\n",
       "  ECa_EMg_2002-2016_rel_sen_slp ECaX_EMgX_2002-2016_rel_sen_slp  \\\n",
       "0                    -0.0102527                     -0.00992399   \n",
       "1                    -0.0101394                      -0.0100083   \n",
       "2                   -0.00722479                     -0.00720022   \n",
       "3                   0.000580972                     0.000841425   \n",
       "4                     -0.012619                     -0.00857364   \n",
       "\n",
       "  ANC_2002-2016_rel_sen_slp  \n",
       "0                 0.0247881  \n",
       "1                 0.0166807  \n",
       "2                 0.0106598  \n",
       "3                 0.0142161  \n",
       "4                  0.017423  \n",
       "\n",
       "[5 rows x 244 columns]"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "del df['data_period']\n",
    "\n",
    "# Melt to \"long\" format\n",
    "melt_df = pd.melt(df, \n",
    "                  id_vars=['project_id', 'project_name', 'station_id', 'station_code', \n",
    "                           'station_name', 'nfc_code', 'type', 'continent', 'country', \n",
    "                           'region', 'subregion', 'latitude', 'longitude', \n",
    "                           'analysis_period', 'par_id', 'include'],\n",
    "                  var_name='stat')\n",
    "\n",
    "# Get only values where include='yes'\n",
    "melt_df = melt_df.query('include == \"yes\"')\n",
    "del melt_df['include']\n",
    "\n",
    "# Build multi-index on everything except \"value\"\n",
    "melt_df.set_index(['project_id', 'project_name', 'station_id', 'station_code', \n",
    "                   'station_name', 'nfc_code', 'type', 'continent', 'country', \n",
    "                   'region', 'subregion', 'latitude', 'longitude', 'par_id', \n",
    "                   'analysis_period', \n",
    "                   'stat'], inplace=True)\n",
    "\n",
    "# Unstack levels of interest to columns\n",
    "wide_df = melt_df.unstack(level=['par_id', 'analysis_period', 'stat'])\n",
    "\n",
    "# Drop unwanted \"value\" level in index\n",
    "wide_df.columns = wide_df.columns.droplevel(0)\n",
    "\n",
    "# Replace multi-index with separate components concatenated with '_'\n",
    "wide_df.columns = [\"_\".join(item) for item in wide_df.columns]\n",
    "\n",
    "# Reset multiindex on rows\n",
    "wide_df = wide_df.reset_index()\n",
    "\n",
    "# Save output\n",
    "out_path = os.path.join(res_fold, 'toc_trends_wide_format.csv')\n",
    "wide_df.to_csv(out_path, index=False, encoding='utf-8')\n",
    "\n",
    "wide_df.head()"
   ]
  }
 ],
 "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": "3.6.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}