Micro-BarChart-Arrays.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Barplot matrices of summer water quality testing\n",
    "\n",
    "1. Summer 2016 and 2017\n",
    "2. Joint project citizen science Hackuarium/hammerdirt\n",
    "\n",
    "### Usage:\n",
    "\n",
    "1. Graphics for citizen science article\n",
    "2. Output for web application\n",
    "3. You may use give us the credit\n",
    "\n",
    "### Contact and further information\n",
    "\n",
    "1. roger@hammerdirt.ch\n",
    "2. https://mwshovel.pythonanywhere.com/dirt/microbiology.html\n",
    "3. http://wiki.hackuarium.ch/w/Main_Page\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib\n",
    "import matplotlib.pyplot as plt\n",
    "import json\n",
    "import re\n",
    "import seaborn\n",
    "import textwrap\n",
    "import matplotlib.patches as mpatches"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Get the 2017 data:\n",
    "\n",
    "#### We are working from the original csv\n",
    "\n",
    "We will need to do the following:\n",
    "\n",
    "1. Dates formatted\n",
    "2. Duplicates removed\n",
    "3. Errors corrected\n",
    "\n",
    "#### see the workbook \"Initial sorting and cat herding\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "a = pd.read_csv('data/2017_Data.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Date                                                           26.06.17\n",
       "Location                                                      Echandens\n",
       "medium                                                         easy_gel\n",
       "Samples                                                               1\n",
       "Sampling_Notes                                              No comments\n",
       "Water_temp                                                           18\n",
       "Plating_notes                                    plate one is Easy Gel \n",
       "Temp_incubation                                                      37\n",
       "P1_qty_sample                                                         4\n",
       "Image_24h_fluo_plate_one                            Filter-16_55_45.JPG\n",
       "P1_fluo_halo_colonies                                                 0\n",
       "P1_fluo_other                                                         0\n",
       "Plate_one_24h_image                                 filter-16_55_53.JPG\n",
       "P1_24h_big_blue                                                       0\n",
       "P1_24h_med_blue                                                       0\n",
       "P1_24h_green                                                          0\n",
       "P1_24h_turq                                                           0\n",
       "P1_24h_pink                                                           0\n",
       "P1_24h_other                                                          0\n",
       "Comments_p1_24h                                             No comments\n",
       "Plate_one_48h_image                                                none\n",
       "P1_48h_big_blue                                                       0\n",
       "P1_48h_med_blue                                                       0\n",
       "P1_48h_green                                                          0\n",
       "P1_48h_turq                                                           0\n",
       "P1_48h_pink                                                           0\n",
       "P1_48h_other                                                          0\n",
       "Comments_p1_48h                                             No comments\n",
       "P2_qty_sample                                                      none\n",
       "Image_24h_fluo_plate_two                                           none\n",
       "                                                ...                    \n",
       "P2_48h_big_blue                                                       0\n",
       "P2_48h_med_blue                                                       0\n",
       "P2_48h_green                                                          0\n",
       "P2_48h_turq                                                           0\n",
       "P2_48h_pink                                                           0\n",
       "P2_48h_other                                                          0\n",
       "Comments_p2_48h               Plate two is non selective \"UNIL\" Kitchen\n",
       "P3_qty_sample                                                         4\n",
       "Image_24h_fluo_plate_three                                         none\n",
       "p3_fluo_halo_colonies                                                 0\n",
       "p3_fluo_other                                                         0\n",
       "Plate_three_24h_image                                              none\n",
       "P3_24h_big_blue                                                       0\n",
       "P3_24h_med_blue                                                       0\n",
       "P3_24h_green                                                          0\n",
       "P3_24h_turq                                                           0\n",
       "P3_24h_pink                                                           0\n",
       "P3_24h_other                                                          0\n",
       "Comments_p3_24h                                             No comments\n",
       "Plate_three_48h_image                                              none\n",
       "P3_48h_big_blue                                                       0\n",
       "P3_48h_med_blue                                                       0\n",
       "P3_48h_green                                                          0\n",
       "P3_48h_turq                                                           0\n",
       "P3_48h_pink                                                           0\n",
       "P3_48h_other                                                          0\n",
       "Comments_p3_48h                                             No comments\n",
       "latitude                                                        46.5347\n",
       "longitude                                                        6.5396\n",
       "Location_Image                                                     none\n",
       "Name: 12, Length: 71, dtype: object"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "a.iloc[12]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# this is a duplicate value in the DF, the data was moved and then added as a record)\n",
    "# so we need to get rid of that:\n",
    "a.drop(12, inplace=True)\n",
    "# there is a column name that doesn't match the others\n",
    "a.rename(columns={'p3_fluo_halo_colonies':'P3_fluo_halo_colonies'}, inplace=True)\n",
    "# there is a sample number for a record where no samples weere taken, we need to fix that:\n",
    "a.loc[(a.Location == 'MRD') & (a.Date == '24.07.17'), \"Samples\"] = 0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "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>Date</th>\n",
       "      <th>Location</th>\n",
       "      <th>medium</th>\n",
       "      <th>Samples</th>\n",
       "      <th>Sampling_Notes</th>\n",
       "      <th>Water_temp</th>\n",
       "      <th>Plating_notes</th>\n",
       "      <th>Temp_incubation</th>\n",
       "      <th>P1_qty_sample</th>\n",
       "      <th>Image_24h_fluo_plate_one</th>\n",
       "      <th>...</th>\n",
       "      <th>P3_48h_big_blue</th>\n",
       "      <th>P3_48h_med_blue</th>\n",
       "      <th>P3_48h_green</th>\n",
       "      <th>P3_48h_turq</th>\n",
       "      <th>P3_48h_pink</th>\n",
       "      <th>P3_48h_other</th>\n",
       "      <th>Comments_p3_48h</th>\n",
       "      <th>latitude</th>\n",
       "      <th>longitude</th>\n",
       "      <th>Location_Image</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>45</th>\n",
       "      <td>24.07.17</td>\n",
       "      <td>MRD</td>\n",
       "      <td>easy_gel</td>\n",
       "      <td>0</td>\n",
       "      <td>Waves at sight, sampling not effective</td>\n",
       "      <td>0</td>\n",
       "      <td>No comments</td>\n",
       "      <td>37</td>\n",
       "      <td>4.0</td>\n",
       "      <td>none</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>No comments</td>\n",
       "      <td>46.431662</td>\n",
       "      <td>6.908962</td>\n",
       "      <td>1500895656638.jpg</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1 rows × 71 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        Date Location    medium  Samples  \\\n",
       "45  24.07.17      MRD  easy_gel        0   \n",
       "\n",
       "                            Sampling_Notes  Water_temp Plating_notes  \\\n",
       "45  Waves at sight, sampling not effective           0   No comments   \n",
       "\n",
       "    Temp_incubation  P1_qty_sample Image_24h_fluo_plate_one  \\\n",
       "45               37            4.0                     none   \n",
       "\n",
       "          ...          P3_48h_big_blue  P3_48h_med_blue P3_48h_green  \\\n",
       "45        ...                      0.0              0.0          0.0   \n",
       "\n",
       "    P3_48h_turq  P3_48h_pink  P3_48h_other  Comments_p3_48h   latitude  \\\n",
       "45          0.0          0.0           0.0      No comments  46.431662   \n",
       "\n",
       "    longitude     Location_Image  \n",
       "45   6.908962  1500895656638.jpg  \n",
       "\n",
       "[1 rows x 71 columns]"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "a.loc[(a.Location == 'MRD') & (a.Date == '24.07.17')]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "# b = pd.read_csv('data/2017_Data/rain2017.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Create the categories/criteria for analysis/comparison\n",
    "\n",
    "1. Create labels for chart output\n",
    "2. Link column names to chart labels\n",
    "2. Create lists of unique values for categories/criteria"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Get the unique dates:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['12.06.17',\n",
       " '19.06.17',\n",
       " '26.06.17',\n",
       " '03.07.17',\n",
       " '10.07.17',\n",
       " '17.07.17',\n",
       " '24.07.17',\n",
       " '31.07.17']"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dates = list(a.Date.unique())\n",
    "dates"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Get the locations of interest:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['VNX', 'SVT', 'MRD']"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "locations = ['VNX', 'SVT',  'MRD']\n",
    "locations"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Get the culture media:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['easy_gel', 'micrology_card', 'unil_kitchen', 'levine']"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mediums = list(a['medium'].unique().copy())\n",
    "mediums"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Get the 24hour results:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "twentyFourHour = ['Date', 'Location', 'medium', 'Samples', 'Sampling_Notes', 'Water_temp', 'Plating_notes', 'Temp_incubation', 'P1_qty_sample', 'P1_fluo_halo_colonies', 'Plate_one_24h_image', 'P1_24h_big_blue', 'P1_24h_med_blue', 'P1_24h_green', 'P1_24h_turq', 'P1_24h_pink', 'P1_24h_other', 'Comments_p1_24h','P2_qty_sample','P2_fluo_halo_colonies', 'Plate_two_24h_image', 'P2_24h_big_blue', 'P2_24h_med_blue', 'P2_24h_green', 'P2_24h_turq', 'P2_24h_pink', 'P2_24h_other', 'Comments_p2_24h', 'P3_qty_sample', 'P3_fluo_halo_colonies', 'Plate_three_24h_image', 'P3_24h_big_blue', 'P3_24h_med_blue', 'P3_24h_green', 'P3_24h_turq', 'P3_24h_pink', 'P3_24h_other', 'Comments_p3_24h']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "twntyFor2017 = a[twentyFourHour]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Get the 48hour results:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "fortyEightHour = ['Date','Location','medium','Samples','Sampling_Notes','Water_temp','Plating_notes','Temp_incubation','P1_qty_sample','P1_fluo_halo_colonies','Plate_one_48h_image','P1_48h_big_blue','P1_48h_med_blue','P1_48h_green','P1_48h_turq','P1_48h_pink','P1_48h_other','Comments_p1_48h','P2_qty_sample','P2_fluo_halo_colonies', 'Plate_two_48h_image','P2_48h_big_blue','P2_48h_med_blue','P2_48h_green', 'P2_48h_turq', 'P2_48h_pink','P2_48h_other','Comments_p2_48h','P3_qty_sample','P3_fluo_halo_colonies','Plate_three_48h_image','P3_48h_big_blue','P3_48h_med_blue', 'P3_48h_green','P3_48h_turq', 'P3_48h_pink','P3_48h_other', 'Comments_p3_48h']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "frtyEght2017 = a[fortyEightHour]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Create week labels and categories"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "weekLabels = ['Week one', 'Week two', 'Week three', 'Week four', 'Week five', 'Week six', 'Week seven', 'Week eight']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "# map the labels to the weeks\n",
    "# we will be comparing the weekly results from one year to the next\n",
    "# this will allow us to call results by week number\n",
    "weeksDates = dict(zip(weekLabels, dates))\n",
    "# inverse this relationship \n",
    "datesWeeks = {value:key for key, value in weeksDates.items()}"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Example: A dictionary that has the week number as key and the date of the sample as value"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'19.06.17'"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "weeksDates[\"Week two\"]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Create labels for colony colors\n",
    "\n",
    "1. Make a dictionary that ties column name to color\n",
    "2. Eliminate the plate reference\n",
    "3. Used to label charts"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "# create keys for charting labels\n",
    "platesColors = ['P1_fluo_halo_colonies','P1_48h_big_blue','P1_48h_med_blue','P1_48h_green','P1_48h_turq','P1_48h_pink','P1_48h_other','P2_fluo_halo_colonies', 'P2_48h_big_blue','P2_48h_med_blue','P2_48h_green', 'P2_48h_turq', 'P2_48h_pink','P2_48h_other','P3_fluo_halo_colonies','P3_48h_big_blue','P3_48h_med_blue', 'P3_48h_green','P3_48h_turq', 'P3_48h_pink','P3_48h_other','P1_fluo_halo_colonies','P1_24h_big_blue', 'P1_24h_med_blue', 'P1_24h_green', 'P1_24h_turq', 'P1_24h_pink', 'P1_24h_other','P2_fluo_halo_colonies','P2_24h_big_blue', 'P2_24h_med_blue', 'P2_24h_green', 'P2_24h_turq', 'P2_24h_pink', 'P2_24h_other','P3_fluo_halo_colonies','P3_24h_big_blue', 'P3_24h_med_blue', 'P3_24h_green', 'P3_24h_turq', 'P3_24h_pink', 'P3_24h_other']\n",
    "\n",
    "def makeKeys(a):\n",
    "    b = re.compile('fluo_halo', re.IGNORECASE)\n",
    "    c={}\n",
    "    for d,e in enumerate(a):\n",
    "        if b.search(a[d]):\n",
    "            f = 'UV Fluo'\n",
    "            g = {a[d]:f}\n",
    "            c.update(g)\n",
    "        else:\n",
    "            f = a[d][7:]\n",
    "            g = {a[d]:f}\n",
    "            c.update(g)\n",
    "    return c\n",
    "            \n",
    "platesColorsKeys = makeKeys(platesColors) "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Example: Feed the function a column name and it will return an abbreviated identifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'med_blue'"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "platesColorsKeys['P1_48h_med_blue']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Group plates by number and species\n",
    "\n",
    "1. We need to get the average of results from three different plates for the same species\n",
    "2. This needs to be iterable and indexable"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Group the plates by plate number"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "pOne48 = ['P1_fluo_halo_colonies','P1_48h_big_blue','P1_48h_med_blue','P1_48h_green',\n",
    "            'P1_48h_turq','P1_48h_pink','P1_48h_other']\n",
    "pTwo48= ['P2_fluo_halo_colonies','P2_48h_big_blue','P2_48h_med_blue',\n",
    "            'P2_48h_green','P2_48h_turq','P2_48h_pink','P2_48h_other']\n",
    "pThree48 = ['P3_fluo_halo_colonies','P3_48h_big_blue','P3_48h_med_blue',\n",
    "              'P3_48h_green','P3_48h_turq','P3_48h_pink','P3_48h_other']\n",
    "pOne24 = ['P1_fluo_halo_colonies', 'P1_24h_big_blue','P1_24h_med_blue',\n",
    "            'P1_24h_green', 'P1_24h_turq', 'P1_24h_pink', 'P1_24h_other']\n",
    "pTwo24 = ['P2_fluo_halo_colonies','P2_24h_big_blue','P2_24h_med_blue',\n",
    "            'P2_24h_green', 'P2_24h_turq', 'P2_24h_pink','P2_24h_other',]\n",
    "pThree24 = ['P3_fluo_halo_colonies', 'P3_24h_big_blue', 'P3_24h_med_blue',\n",
    "              'P3_24h_green', 'P3_24h_turq', 'P3_24h_pink','P3_24h_other']\n",
    "# Group the plates by incubation time\n",
    "platesFortyEight = [pOne48, pTwo48, pThree48]\n",
    "platesTwentyFour = [pOne24, pTwo24, pThree24]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[['P1_fluo_halo_colonies',\n",
       "  'P1_24h_big_blue',\n",
       "  'P1_24h_med_blue',\n",
       "  'P1_24h_green',\n",
       "  'P1_24h_turq',\n",
       "  'P1_24h_pink',\n",
       "  'P1_24h_other'],\n",
       " ['P2_fluo_halo_colonies',\n",
       "  'P2_24h_big_blue',\n",
       "  'P2_24h_med_blue',\n",
       "  'P2_24h_green',\n",
       "  'P2_24h_turq',\n",
       "  'P2_24h_pink',\n",
       "  'P2_24h_other'],\n",
       " ['P3_fluo_halo_colonies',\n",
       "  'P3_24h_big_blue',\n",
       "  'P3_24h_med_blue',\n",
       "  'P3_24h_green',\n",
       "  'P3_24h_turq',\n",
       "  'P3_24h_pink',\n",
       "  'P3_24h_other']]"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "platesTwentyFour"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Group the species (color) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "def makeGroups(a):\n",
    "    b = np.arange(len(a[0]))\n",
    "    c=[]\n",
    "    for d in b:\n",
    "        e = []\n",
    "        for f in a:\n",
    "            e.append(f[d])\n",
    "        c.append(e)\n",
    "    return c\n",
    "fortyEightPlates = makeGroups(platesFortyEight)\n",
    "twentyFourPlates = makeGroups(platesTwentyFour)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Example: the output should be all the plate counts for one color\n",
    "\n",
    "1. We use this to call the columns on a given day\n",
    "2. Then average the results for the three columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['P1_24h_green', 'P2_24h_green', 'P3_24h_green']"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "twentyFourPlates[3]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Format data for charting and JSON output\n",
    "\n",
    "1. Use the location list to select records from one location\n",
    "2. Use the mediums list to select results based on the growth medium used\n",
    "3. Use the weeks list to collect the results per growth medium per week\n",
    "4. Pass those results on to other functions that calculate and select specific values"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Limit the dataframe to the location and culture medium desired"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "# functions that reduce the Data set to a specific location and medium\n",
    "# Returns a DataFrame where each row is a unique date\n",
    "\n",
    "def getLocation(df, places):\n",
    "    b = df[df.Location == places]\n",
    "    return b\n",
    "def getMedium(df, medium):\n",
    "    b = df[df.medium == medium]\n",
    "    return b\n",
    "def getWeeklyValues(df, week):\n",
    "    d = []\n",
    "    for a,b in enumerate(week):\n",
    "        c = df[df.Date == week[a]]\n",
    "        d.append(c)\n",
    "    return d\n",
    "# This function calls the other functions to produce the desired output\n",
    "# a list of dataframes, each dataFrame is one row of data that contains the values\n",
    "# for a specified day. location and culture medium\n",
    "def getItAll(df, places, medium, week):\n",
    "    a = getLocation(df, places)\n",
    "    c = getMedium(a, medium)\n",
    "    d = getWeeklyValues(c, week)\n",
    "    return d\n",
    "\n",
    "x = getItAll(a, locations[2], mediums[0], dates)        "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Example: call the variable \"x\" and give it an index number\n",
    "\n",
    "1. You can use column indexing by name"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "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>Date</th>\n",
       "      <th>Location</th>\n",
       "      <th>medium</th>\n",
       "      <th>Sampling_Notes</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>45</th>\n",
       "      <td>24.07.17</td>\n",
       "      <td>MRD</td>\n",
       "      <td>easy_gel</td>\n",
       "      <td>Waves at sight, sampling not effective</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        Date Location    medium                          Sampling_Notes\n",
       "45  24.07.17      MRD  easy_gel  Waves at sight, sampling not effective"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# here we are just calling the columns that represent the sorting criteria\n",
    "# there are 72 columns in all\n",
    "# \"x\" will be passed on to the next set of functions\n",
    "x[6][[\"Date\", \"Location\", \"medium\",\"Sampling_Notes\"]]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Use the results from the above functions and retrieve key information and perform calculations: \n",
    "\n",
    "1. Retrieve comments from sampling and counting\n",
    "2. Get the average color/day/location \n",
    "3. Identify \"Below detectable limit\" and \"Too many to count\"\n",
    "4. Put that all in a dictionary to be exported in JSON format"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Getting comments:\n",
    "\n",
    "# The sampling comments are in one column\n",
    "# This information was entered by the person in the field who took the sample\n",
    "# If there are no samples \"Samples.values = 0\" then there should be a comment\n",
    "# If there are no samples and no comments we fill it in with \"Samples not taken\"\n",
    "# If there are samples then no comments would be registered\n",
    "\n",
    "def getSamplingComments(a):    \n",
    "    f = {}\n",
    "    for b in a:\n",
    "        c = b[\"Date\"].values[0]\n",
    "        d = b['Samples'].values[0]\n",
    "        if d == 0:\n",
    "            e = b[\"Sampling_Notes\"].item()\n",
    "            if e == \"No comments\":\n",
    "                no_t = 'Samples not taken'\n",
    "                f.update({d:no_t})\n",
    "            else:\n",
    "                no_t = e\n",
    "                f.update({c:no_t})                \n",
    "        else:\n",
    "            no_t = \"none\"\n",
    "            f.update({c:no_t})\n",
    "    return f\n",
    "\n",
    "# The count comments were provided by the person doing the colony counts\n",
    "# Specifically we wanted to know if the value was \"Too many to count\" -- tmtc\n",
    "# Therefore there is one comment block for each plate\n",
    "# We search each comment block for \"tmtc\"\n",
    "# if we find tmtc then the value is stored in a dictionary {Date:Comment}\n",
    "\n",
    "def getCountComments(a):\n",
    "    p = re.compile('tmtc', re.IGNORECASE)\n",
    "    countComments = ['Date','Location', 'Comments_p1_24h', 'Comments_p1_48h', 'Comments_p2_24h','Comments_p2_48h', 'Comments_p3_24h', 'Comments_p3_48h'] \n",
    "    b = {}\n",
    "    for a in x:\n",
    "        c = a[countComments]\n",
    "        e = c[\"Date\"].item()\n",
    "        f = {e:[]}\n",
    "        for d in countComments:           \n",
    "            if p.search(a[d].item()):                \n",
    "                g = c[d].item()\n",
    "                f[e].append(g)#               \n",
    "            else:\n",
    "                pass\n",
    "        b.update(f)\n",
    "    return b\n",
    "          \n",
    "          \n",
    "# The number of samples is noted in the dataframe\n",
    "# get the value and store it in a dictionary\n",
    "def getNumberOfSamples(a):\n",
    "    f = {}\n",
    "    for b in a:\n",
    "        c = b.Samples.item()\n",
    "        d = b.Date.item()\n",
    "        e = {d:c}\n",
    "        f.update(e)\n",
    "    return f\n",
    "\n",
    "# Get the average per color group, per day\n",
    "# Call the species groups ie.. fortyEightPlates\n",
    "# Iterate through the species groups\n",
    "# Averge the values greater than zero\n",
    "# Identify those values = zero and store in a dictionary \"Below detectable limit\"\n",
    "# Colors with BDL are given a nominal value of 1 (that way there will be a hint of color on the chart)\n",
    "# Store all results in a dicitionary where the keys are the unique dates\n",
    "\n",
    "def getAverage(plates,data):\n",
    "    c = np.arange(len(plates))\n",
    "    d = np.arange(len(data))\n",
    "    e = {0.5:200, 1:100, 4:25}\n",
    "    f = {}\n",
    "    for g in d:\n",
    "        h = data[g][\"Date\"].item()\n",
    "        i = data[g]['P1_qty_sample'].item()\n",
    "        r = data[g][\"Location\"].item()\n",
    "        j = e[i]\n",
    "        k = {h:{}}\n",
    "        o = {\"BDL\":[]}\n",
    "        s = {\"Qty plated\":i}\n",
    "        aves = {\"Averages\":[]}\n",
    "        for l in c:\n",
    "            m = data[g][plates[l]].mean(axis=1).item()\n",
    "            p = platesColorsKeys[plates[l][0]]\n",
    "            ave = m*j            \n",
    "            if ave == 0:\n",
    "                o[\"BDL\"].append(p)\n",
    "                ave += 1\n",
    "            aves[\"Averages\"].append(ave)        \n",
    "        k[h].update(aves)\n",
    "        k[h].update(o)\n",
    "        k[h].update(s)\n",
    "        f.update(k)\n",
    "    q = {r:f}\n",
    "    return q\n",
    "\n",
    "# A function for converting the dates to week numbers\n",
    "def dateToWeek(a):\n",
    "    f = {}\n",
    "    for b,c in a.items():\n",
    "        d = datesWeeks[b]\n",
    "        e = {d:c}\n",
    "        f.update(e)\n",
    "    return f  "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Example: Plating comments:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'12.06.17': 'none',\n",
       " '19.06.17': 'none',\n",
       " '26.06.17': 'none',\n",
       " '03.07.17': 'none',\n",
       " '10.07.17': 'none',\n",
       " '17.07.17': 'none',\n",
       " '24.07.17': 'Waves at sight, sampling not effective',\n",
       " '31.07.17': 'none'}"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "theComs = getSamplingComments(x)\n",
    "theComs"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Example: Counting comments"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'12.06.17': [],\n",
       " '19.06.17': [],\n",
       " '26.06.17': [],\n",
       " '03.07.17': [],\n",
       " '10.07.17': [],\n",
       " '17.07.17': ['TMTC'],\n",
       " '24.07.17': [],\n",
       " '31.07.17': []}"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "countingComments = getCountComments(x)\n",
    "countingComments"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Example: Number of samples"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'12.06.17': 3,\n",
       " '19.06.17': 3,\n",
       " '26.06.17': 3,\n",
       " '03.07.17': 3,\n",
       " '10.07.17': 3,\n",
       " '17.07.17': 3,\n",
       " '24.07.17': 0,\n",
       " '31.07.17': 3}"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "numberOfSamples = getNumberOfSamples(x)\n",
    "numberOfSamples"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "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>Date</th>\n",
       "      <th>Location</th>\n",
       "      <th>medium</th>\n",
       "      <th>Samples</th>\n",
       "      <th>Sampling_Notes</th>\n",
       "      <th>Water_temp</th>\n",
       "      <th>Plating_notes</th>\n",
       "      <th>Temp_incubation</th>\n",
       "      <th>P1_qty_sample</th>\n",
       "      <th>Image_24h_fluo_plate_one</th>\n",
       "      <th>...</th>\n",
       "      <th>P3_48h_big_blue</th>\n",
       "      <th>P3_48h_med_blue</th>\n",
       "      <th>P3_48h_green</th>\n",
       "      <th>P3_48h_turq</th>\n",
       "      <th>P3_48h_pink</th>\n",
       "      <th>P3_48h_other</th>\n",
       "      <th>Comments_p3_48h</th>\n",
       "      <th>latitude</th>\n",
       "      <th>longitude</th>\n",
       "      <th>Location_Image</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>45</th>\n",
       "      <td>24.07.17</td>\n",
       "      <td>MRD</td>\n",
       "      <td>easy_gel</td>\n",
       "      <td>0</td>\n",
       "      <td>Waves at sight, sampling not effective</td>\n",
       "      <td>0</td>\n",
       "      <td>No comments</td>\n",
       "      <td>37</td>\n",
       "      <td>4.0</td>\n",
       "      <td>none</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>No comments</td>\n",
       "      <td>46.431662</td>\n",
       "      <td>6.908962</td>\n",
       "      <td>1500895656638.jpg</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1 rows × 71 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        Date Location    medium  Samples  \\\n",
       "45  24.07.17      MRD  easy_gel        0   \n",
       "\n",
       "                            Sampling_Notes  Water_temp Plating_notes  \\\n",
       "45  Waves at sight, sampling not effective           0   No comments   \n",
       "\n",
       "    Temp_incubation  P1_qty_sample Image_24h_fluo_plate_one  \\\n",
       "45               37            4.0                     none   \n",
       "\n",
       "          ...          P3_48h_big_blue  P3_48h_med_blue P3_48h_green  \\\n",
       "45        ...                      0.0              0.0          0.0   \n",
       "\n",
       "    P3_48h_turq  P3_48h_pink  P3_48h_other  Comments_p3_48h   latitude  \\\n",
       "45          0.0          0.0           0.0      No comments  46.431662   \n",
       "\n",
       "    longitude     Location_Image  \n",
       "45   6.908962  1500895656638.jpg  \n",
       "\n",
       "[1 rows x 71 columns]"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x[6]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "8"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(x)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Example: Daily averages "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'Averages': [1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0],\n",
       " 'BDL': ['UV Fluo', 'big_blue', 'med_blue', 'green', 'turq', 'pink', 'other'],\n",
       " 'Qty plated': 4.0}"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "theResults = getAverage(fortyEightPlates, x)\n",
    "theResults['MRD']['12.06.17']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Putting it all together\n",
    "\n",
    "1. Combine all the results for each location and sample day\n",
    "2. Use a dictionary format where the key is location\n",
    "3. Create bar-chart array\n",
    "4. Export to JSON create Web output"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Combine all the results and iterate through the locations\n",
    "\n",
    "1. Create one dictionary per culture medium"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "def getAllLocations(df, places, medium, dates):\n",
    "    a = {}\n",
    "    b = np.arange(len(places))\n",
    "    for c in b:\n",
    "        d = getItAll(df, places[c], medium, dates)\n",
    "        e = {places[c]:d}\n",
    "        a.update(e)\n",
    "    return a\n",
    "allResults = getAllLocations(a, locations, mediums[0], dates)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "def makeOutput(data, places, plates):\n",
    "    f = {}\n",
    "    for i,n in enumerate(places):\n",
    "        a = data[n]\n",
    "        b = getAverage(plates, a)\n",
    "        c = getSamplingComments(a)\n",
    "        d = getCountComments(a)\n",
    "        e = getNumberOfSamples(a)\n",
    "        g = list(b[n].keys())\n",
    "        for date  in g:\n",
    "            b[n][date].update({\"Sample Comments\":c[date]})\n",
    "            b[n][date].update({\"Count Comments\":d[date]})\n",
    "            b[n][date].update({\"Number of Samples\":e[date]})          \n",
    "            \n",
    "        f.update(b)\n",
    "    return f\n",
    "theOutput = makeOutput(allResults, locations,twentyFourPlates)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'12.06.17': {'Averages': [1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0],\n",
       "  'BDL': ['UV Fluo', 'big_blue', 'med_blue', 'green', 'turq', 'pink', 'other'],\n",
       "  'Qty plated': 4.0,\n",
       "  'Sample Comments': 'none',\n",
       "  'Count Comments': [],\n",
       "  'Number of Samples': 3},\n",
       " '19.06.17': {'Averages': [1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0],\n",
       "  'BDL': ['UV Fluo', 'big_blue', 'med_blue', 'green', 'turq', 'pink', 'other'],\n",
       "  'Qty plated': 4.0,\n",
       "  'Sample Comments': 'none',\n",
       "  'Count Comments': [],\n",
       "  'Number of Samples': 3},\n",
       " '26.06.17': {'Averages': [1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0],\n",
       "  'BDL': ['UV Fluo', 'big_blue', 'med_blue', 'green', 'turq', 'pink', 'other'],\n",
       "  'Qty plated': 4.0,\n",
       "  'Sample Comments': 'none',\n",
       "  'Count Comments': [],\n",
       "  'Number of Samples': 3},\n",
       " '03.07.17': {'Averages': [91.66666666666666,\n",
       "   91.66666666666666,\n",
       "   83.33333333333334,\n",
       "   1.0,\n",
       "   1.0,\n",
       "   16.666666666666664,\n",
       "   1.0],\n",
       "  'BDL': ['green', 'turq', 'other'],\n",
       "  'Qty plated': 4.0,\n",
       "  'Sample Comments': 'none',\n",
       "  'Count Comments': [],\n",
       "  'Number of Samples': 3},\n",
       " '10.07.17': {'Averages': [1.0,\n",
       "   1.0,\n",
       "   133.33333333333331,\n",
       "   1.0,\n",
       "   1.0,\n",
       "   66.66666666666666,\n",
       "   1.0],\n",
       "  'BDL': ['UV Fluo', 'big_blue', 'green', 'turq', 'other'],\n",
       "  'Qty plated': 1.0,\n",
       "  'Sample Comments': 'none',\n",
       "  'Count Comments': [],\n",
       "  'Number of Samples': 3},\n",
       " '17.07.17': {'Averages': [8.333333333333332,\n",
       "   8.333333333333332,\n",
       "   8.333333333333332,\n",
       "   1.0,\n",
       "   1.0,\n",
       "   1.0,\n",
       "   125.0],\n",
       "  'BDL': ['green', 'turq', 'pink'],\n",
       "  'Qty plated': 4.0,\n",
       "  'Sample Comments': 'none',\n",
       "  'Count Comments': ['TMTC'],\n",
       "  'Number of Samples': 3},\n",
       " '24.07.17': {'Averages': [8.333333333333332,\n",
       "   8.333333333333332,\n",
       "   1.0,\n",
       "   1.0,\n",
       "   1.0,\n",
       "   1.0,\n",
       "   1.0],\n",
       "  'BDL': ['med_blue', 'green', 'turq', 'pink', 'other'],\n",
       "  'Qty plated': 4.0,\n",
       "  'Sample Comments': 'none',\n",
       "  'Count Comments': [],\n",
       "  'Number of Samples': 3},\n",
       " '31.07.17': {'Averages': [16.666666666666664, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0],\n",
       "  'BDL': ['big_blue', 'med_blue', 'green', 'turq', 'pink', 'other'],\n",
       "  'Qty plated': 4.0,\n",
       "  'Sample Comments': 'none',\n",
       "  'Count Comments': [],\n",
       "  'Number of Samples': 3}}"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "theOutput[\"SVT\"] \n",
    "\n",
    "##  2017 Create a graph that displays the average colony counts per day & location \n",
    "# 1. Identify and note \"Below detectable limit\"\n",
    "# 2. Create a matrix Row = sample day, column = location\n",
    "# 3. Note the days and locations samples were not taken\n",
    "# 4. Make output SVG for web applications"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "code_folding": []
   },
   "outputs": [],
   "source": [
    "# start the function\n",
    "def plot_ez_totals(x, bars, svgtitle, pngtitle, title, subtitle, colors, legendColors):\n",
    "    # number of rows depends on the number of weeks\n",
    "    rows = len(dates)\n",
    "    # the number of columns depends on the number of locations\n",
    "    cols = len(x)\n",
    "    # this is the number of bars in each chart\n",
    "    # this depends on the number of color groups\n",
    "    ind = np.arange(bars)#<--- this can be made a variable\n",
    "    # the width of each bar\n",
    "    # width one means there is no space between bars\n",
    "    width = 1\n",
    "    # assign appropriate colors per grop\n",
    "    # note the index number of each color matches the index number of the values from the first fucntion\n",
    "    colss = colors\n",
    "    # set up the figure and the array of charts\n",
    "    f, axar = plt.subplots(rows, cols, figsize=(12,24), sharey='row')\n",
    "    # create some space in between the rows for x axis labels, and titles\n",
    "    f.subplots_adjust(hspace=0.6, wspace=0.1)\n",
    "    # run through the results generated by the previous fucntion\n",
    "    for i in np.arange(cols):\n",
    "        a = i\n",
    "        b = locations[a]\n",
    "        c = x[b]\n",
    "        d = datesWeeks[dates[a]]\n",
    "        for n,day in enumerate(dates):\n",
    "            if i == 0:\n",
    "                if c[day]['Sample Comments'] != \"none\":\n",
    "                    axar[n, i].text(0.05, 10, c[day]['Sample Comments'], fontsize=10)\n",
    "                    axar[n, i].set_title(b)\n",
    "                    axar[n, i].set_ylabel(\"Colony forming units per 100mL\", fontdict={'fontsize': 7, 'fontweight': 'medium'},\n",
    "                                      labelpad=1,)\n",
    "                    axar[n, i].xaxis.set_major_locator(plt.NullLocator())\n",
    "                    axar[n, i].xaxis.set_major_formatter(plt.NullFormatter())\n",
    "                else:\n",
    "                    axar[n, i].bar(ind, c[day][\"Averages\"], width, color=colss)\n",
    "                    axar[n, i].set_ylim(0, max(c[day][\"Averages\"]) + 50)\n",
    "                    axar[n, i].set_title(b + ': ' +  day + ', samples: ' + str(c[day]['Number of Samples']) + ', ' +\n",
    "                                     str(c[day]['Qty plated'])+ 'mL', fontdict={'fontsize': 10, 'fontweight': 'medium'})\n",
    "                    labels =  \", \".join(c[day][\"BDL\"])\n",
    "                    axis_label = textwrap.fill(labels, width=30)\n",
    "                    axar[n, i].set_xlabel('BDL : '+ str(axis_label))\n",
    "                    axar[n, i].set_ylabel(\"Colony forming units per 100mL\", fontdict={'fontsize': 7, 'fontweight': 'medium'},\n",
    "                                      labelpad=1,)\n",
    "                    axar[n, i].xaxis.set_major_locator(plt.NullLocator())\n",
    "                    axar[n, i].xaxis.set_major_formatter(plt.NullFormatter())\n",
    "            else:\n",
    "                if c[day]['Sample Comments'] != \"none\":\n",
    "                    axar[n, i].text(0.05, 10, c[day]['Sample Comments'], fontsize=10)\n",
    "                    axar[n, i].set_title(b)\n",
    "                    axar[n, i].xaxis.set_major_locator(plt.NullLocator())\n",
    "                    axar[n, i].xaxis.set_major_formatter(plt.NullFormatter())\n",
    "                else:\n",
    "                    axar[n, i].bar(ind, c[day][\"Averages\"], width, color=colss)\n",
    "                    axar[n, i].set_ylim(0, max(c[day][\"Averages\"]) + 50)\n",
    "                    axar[n, i].set_title(b + ': ' +  day + ', samples: ' + str(c[day]['Number of Samples']) + ', ' +\n",
    "                                     str(c[day]['Qty plated'])+ 'mL', fontdict={'fontsize': 10, 'fontweight': 'medium'})\n",
    "                    labels =  \", \".join(c[day][\"BDL\"])\n",
    "                    axis_label = textwrap.fill(labels, width=30)\n",
    "                    axar[n, i].set_xlabel('BDL : '+ str(axis_label))                    \n",
    "                    axar[n, i].xaxis.set_major_locator(plt.NullLocator())\n",
    "                    axar[n, i].xaxis.set_major_formatter(plt.NullFormatter())\n",
    "                \n",
    "    \n",
    "    plt.annotate(title, xy=(0.5, .98),  xycoords=\"figure fraction\",                 \n",
    "                 size=18, ha='center', va=\"top\")\n",
    "    plt.annotate(subtitle, xy=(0.5, .96),  xycoords=\"figure fraction\",                 \n",
    "                 size=12, ha='center', va=\"top\")\n",
    "   \n",
    "    def makePatches(colors, legendColors):\n",
    "        a = []\n",
    "        for i,b in enumerate(colors):\n",
    "            c = mpatches.Patch(color=colors[i], label=legendColors[i])\n",
    "            a.append(c)\n",
    "        return a\n",
    "        \n",
    "    plt.legend(handles=makePatches(colors, legendColors),bbox_to_anchor=(0.11,.87, .8, .08),\n",
    "               mode=\"expand\",loc=\"center\", ncol=len(colors), bbox_transform=f.transFigure )\n",
    "    plt.savefig(svgtitle)\n",
    "    plt.savefig(pngtitle)\n",
    "    plt.show()\n",
    "    plt.close()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Define the variables for the charting function\n",
    "\n",
    "1. Correlate species to color\n",
    "2. Identify the colors you are going to use\n",
    "3. Make a title\n",
    "4. Make a subtitle"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "speciesColor = ['E.coli UV-fluo','E.coli','Other coliforms', 'Salmonella','Salmonella','Aeromonas','other',]\n",
    "colors = ['darkviolet','darkblue', 'dodgerblue', 'green', 'turquoise', 'pink', 'lightslategray']\n",
    "title = \"Average 24 hour colony counts June-July 2017: ECA Check Plus UV\"\n",
    "subTitle = \"MRD = Montreux marché, SVT = Sauvetage, VNX = Vernex, BDL = Below detectable limit\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Call the chart function with the defined variables"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x1728 with 24 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_ez_totals(theOutput, 7, \"data/charts/2017_BARCHART_ARRAY.svg\", \"data/charts/2017_BARCHART_ARRAY.png\", title, subTitle, colors, speciesColor)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Adjust for 2016 data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Get the data:\n",
    "\n",
    "#### This data has already been transformed\n",
    "\n",
    "1. Dates formatted\n",
    "2. Duplicates removed\n",
    "3. Errors corrected\n",
    "\n",
    "#### see the workbook \"Initial sorting and cat herding\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "c = pd.read_csv('data/2016_Data.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Create the categories/cirteria for analysis/comparison\n",
    "\n",
    "1. Create labels for chart output\n",
    "2. Link column names to chart labels\n",
    "2. Create lists of unique values for categories/criteria"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Get the unique dates:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['2016-06-21',\n",
       " '2016-06-28',\n",
       " '2016-07-05',\n",
       " '2016-07-12',\n",
       " '2016-07-19',\n",
       " '2016-07-26',\n",
       " '2016-08-02',\n",
       " '2016-08-09']"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dates16 = list(c.Date.unique())\n",
    "dates16.sort()\n",
    "dates16"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Get the locations of interest:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['VNX', 'SVT', 'MRD']"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "locations"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Get the culture mediums:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['easy_gel']"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "medium16 = ['easy_gel']\n",
    "medium16"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Get the 24hour results:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [],
   "source": [
    "twntyFor2016 = c"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Get the 48hour results:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [],
   "source": [
    "# none"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Create week labels and categories"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [],
   "source": [
    "weeksDates16 = dict(zip(weekLabels, dates16))\n",
    "# inverse this relationship \n",
    "datesWeeks16 = {value:key for key, value in weeksDates16.items()}"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Create labels for colony colors\n",
    "\n",
    "1. Make a dictionary that ties column name to color\n",
    "2. Eliminate the plate reference\n",
    "3. Used to label charts"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [],
   "source": [
    "platesColors16 = ['P1_24h_big_blue', 'P1_24h_med_blue','P1_24h_other', 'P1_24h_pink', 'P1_24h_turq', 'P1_qty_sample',\n",
    "                  'P2_24h_big_blue', 'P2_24h_med_blue', 'P2_24h_other', 'P2_24h_pink','P2_24h_turq', 'P3_24h_big_blue', 'P3_24h_med_blue', 'P3_24h_other',\n",
    "                  'P3_24h_pink', 'P3_24h_turq']\n",
    "platesColorsKeys16 = makeKeys(platesColors16) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'other'"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "platesColorsKeys16['P1_24h_other']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "heading_collapsed": true
   },
   "source": [
    "#### Group plates by number and species\n",
    "\n",
    "1. We need to get the average of results from three different plates for the same species\n",
    "2. This needs to be iterable and indexable"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Group the plates by plate number"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [],
   "source": [
    "pOne24 = ['P1_24h_big_blue', 'P1_24h_med_blue', 'P1_24h_turq', 'P1_24h_pink', 'P1_24h_other']\n",
    "pTwo24 = ['P2_24h_big_blue', 'P2_24h_med_blue', 'P2_24h_turq', 'P2_24h_pink', 'P2_24h_other']\n",
    "pThree24 = ['P3_24h_big_blue', 'P3_24h_med_blue', 'P3_24h_turq', 'P3_24h_pink', 'P3_24h_other']\n",
    "platesTwentyFour16 = [pOne24, pTwo24, pThree24]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Group the species(color) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [],
   "source": [
    "twentyFourPlates16 = makeGroups(platesTwentyFour16)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Example: the output should be all the plate counts for one color\n",
    "\n",
    "1. We use this to call the columns on a given day\n",
    "2. Then average the results for the three columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['P1_24h_med_blue', 'P2_24h_med_blue', 'P3_24h_med_blue']"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "twentyFourPlates16[1]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Limit the dataframe to the location and culture medium desired\n",
    "\n",
    "1. 2016 data has only one culture medium - \"easy-gel\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [],
   "source": [
    "# functions that reduce the Data set to a specific location \n",
    "# Returns a DataFrame where each row is a unique date\n",
    "\n",
    "def getLocation(df, places):\n",
    "    b = df[df.Location == places]\n",
    "    return b\n",
    "def getWeeklyValues(df, week):\n",
    "    d = []\n",
    "    for a,b in enumerate(week):\n",
    "        c = df[df.Date == week[a]]\n",
    "        d.append(c)\n",
    "    return d\n",
    "# This function calls the other functions to produce the desired output\n",
    "# a list of dataframes, each dataFrame is one row of data that contains the values\n",
    "# for a specified day. location and culture medium\n",
    "def getItAll16(df, places, week):\n",
    "    a = getLocation(df, places)\n",
    "    d = getWeeklyValues(a, week)\n",
    "    return d\n",
    "\n",
    "x16 = getItAll16(c, locations[2], dates16)   "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "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>Date</th>\n",
       "      <th>Location</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2016-06-28</td>\n",
       "      <td>MRD</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         Date Location\n",
       "4  2016-06-28      MRD"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# here we are just calling the columns that represent the sorting criteria\n",
    "# there are 72 columns in all\n",
    "# \"x\" will be passed on to the next set of functions\n",
    "x16[1][[\"Date\", \"Location\"]]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Use the results from the above functions and retrieve key information and perform calculations: \n",
    "\n",
    "1. ~~Retrieve comments from sampling and counting~~\n",
    "2. Get the average color/day/location \n",
    "3. Identify \"Below detectable limit\" and \"Too many to count\"\n",
    "4. Put that all in a dictionary to be exported in JSON format\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [],
   "source": [
    "# In 2016 there were three samples per sampleday\n",
    "# get the value and store it in a dictionary\n",
    "def getNumberOfSamples16(a):\n",
    "    f = {}\n",
    "    for b in a:\n",
    "        c = 3\n",
    "        d = b.Date.item()\n",
    "        e = {d:c}\n",
    "        f.update(e)\n",
    "    return f\n",
    "\n",
    "# Get the average per color group, per day\n",
    "# Call the species groups ie.. fortyEightPlates\n",
    "# Iterate through the species groups\n",
    "# Averge the values greater than zero\n",
    "# Identify those values = zero and store in a dictionary \"Below detecable limit\"\n",
    "# Colors with BDL are given a nominal value of 1 (that way there will be a hint of color on the chart)\n",
    "# Store all results in a dicitionary where the keys are the unique dates\n",
    "\n",
    "def getAverage(plates,data):\n",
    "    c = np.arange(len(plates))\n",
    "    d = np.arange(len(data))\n",
    "    e = {0.4: 250, 0.5:200, 1:100, 4:25}\n",
    "    f = {}\n",
    "    for g in d:\n",
    "        h = data[g][\"Date\"].item()\n",
    "        i = data[g]['P1_qty_sample'].item()\n",
    "        r = data[g][\"Location\"].item()\n",
    "        j = e[i]\n",
    "        k = {h:{}}\n",
    "        o = {\"BDL\":[]}\n",
    "        s = {\"Qty plated\":i}\n",
    "        aves = {\"Averages\":[]}\n",
    "        for l in c:\n",
    "            m = data[g][plates[l]].mean(axis=1).item()\n",
    "            p = platesColorsKeys[plates[l][0]]\n",
    "            ave = m*j            \n",
    "            if ave == 0:\n",
    "                o[\"BDL\"].append(p)\n",
    "                ave += 1\n",
    "            aves[\"Averages\"].append(ave)         \n",
    "        k[h].update(aves)\n",
    "        k[h].update(o)\n",
    "        k[h].update(s)\n",
    "        f.update(k)\n",
    "    q = {r:f}\n",
    "    return q\n",
    "\n",
    "# A function for converting the dates to week numbers\n",
    "def dateToWeek(a):\n",
    "    f = {}\n",
    "    for b,c in a.items():\n",
    "        d = datesWeeks[b]\n",
    "        e = {d:c}\n",
    "        f.update(e)\n",
    "    return f  "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Example: Number of samples"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'2016-06-21': 3,\n",
       " '2016-06-28': 3,\n",
       " '2016-07-05': 3,\n",
       " '2016-07-12': 3,\n",
       " '2016-07-19': 3,\n",
       " '2016-07-26': 3,\n",
       " '2016-08-02': 3,\n",
       " '2016-08-09': 3}"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "numberOfSamples16 = getNumberOfSamples16(x16)\n",
    "numberOfSamples16"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Example: Daily averages "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'Averages': [145.0, 25.0, 125.0], 'BDL': [], 'Qty plated': 4.0}"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "theResults16 = getAverage(platesTwentyFour16 , x16)\n",
    "theResults16['MRD']['2016-06-28']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Putting it all together\n",
    "\n",
    "1. Combine all the results for each location and sample day\n",
    "2. Use a dictionary format where the key is location\n",
    "3. Create bar-chart array\n",
    "4. Export to JSON create Web output"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### combine all the results and iterate through the locations\n",
    "\n",
    "1. Create one dictionary per culture medium"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [],
   "source": [
    "def getAllLocations16(df, places, dates):\n",
    "    a = {}\n",
    "    b = np.arange(len(places))\n",
    "    for c in b:\n",
    "        d = getItAll16(df, places[c], dates)\n",
    "        e = {places[c]:d}\n",
    "        a.update(e)\n",
    "    return a\n",
    "allResults16 = getAllLocations16(c, locations, dates16)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [],
   "source": [
    "def makeOutput16(data, places, plates):\n",
    "    f = {}\n",
    "    for i,n in enumerate(places):\n",
    "        a = data[n]\n",
    "        b = getAverage(plates, a)\n",
    "        e = getNumberOfSamples16(a)\n",
    "        g = list(b[n].keys())\n",
    "        for date  in g:\n",
    "            b[n][date].update({\"Number of Samples\":e[date]})           \n",
    "        f.update(b)\n",
    "    return f\n",
    "theOutput16 = makeOutput16(allResults16, locations,twentyFourPlates16)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'Averages': [66.66666666666666,\n",
       "  466.6666666666667,\n",
       "  100.0,\n",
       "  266.66666666666663,\n",
       "  1.0],\n",
       " 'BDL': ['other'],\n",
       " 'Qty plated': 1.0,\n",
       " 'Number of Samples': 3}"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "theOutput16[\"MRD\"]['2016-06-21']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2016   Create a graph that displays the average colony counts per day & location \n",
    "1. Identify and note \"Below detectable limit\"\n",
    "2. Create a matrix Row = sample day, column = location\n",
    "3. Note the days and locations samples were not taken\n",
    "4. Make output SVG for web applications\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {
    "code_folding": []
   },
   "outputs": [],
   "source": [
    "# start the function\n",
    "def plot_ez_totals16(x, bars, svgtitle, pngtitle, title, subtitle, colors, legendColors):\n",
    "    # number of rows depends on the number of weeks\n",
    "    rows = len(dates16)\n",
    "    # the number of columns depends on the number of locations\n",
    "    cols = len(x)\n",
    "    # this is the number of bars in each chart\n",
    "    # this depends on the number of color groups\n",
    "    ind = np.arange(bars)#<--- this can be made a variable\n",
    "    # the width of each bar\n",
    "    # width one means there is no space between bars\n",
    "    width = 1\n",
    "    # assign appropriate colors per grop\n",
    "    # note the index number of each color matches the index number of the values from the first fucntion\n",
    "    colss = colors\n",
    "    # set up the figure and the array of charts\n",
    "    f, axar = plt.subplots(rows, cols, figsize=(12,24), sharey='row')\n",
    "    # create some space in between the rows for x axis labels, and titles\n",
    "    f.subplots_adjust(hspace=0.6, wspace=0.1)\n",
    "    # run through the results generated by the previous fucntion\n",
    "    for i in np.arange(cols):\n",
    "        a = i\n",
    "        b = locations[a]\n",
    "        c = x[b]\n",
    "        d = datesWeeks16[dates16[a]]\n",
    "        for n,day in enumerate(dates16):\n",
    "            if i == 0:\n",
    "                axar[n, i].bar(ind, c[day][\"Averages\"], width, color=colss)\n",
    "                axar[n, i].set_ylim(0, max(c[day][\"Averages\"]) + 50)\n",
    "                axar[n, i].set_title(b + ': ' +  day + ', samples: ' + str(c[day]['Number of Samples']) + ', ' +\n",
    "                                 str(c[day]['Qty plated'])+ 'mL', fontdict={'fontsize': 10, 'fontweight': 'medium'})\n",
    "                labels =  \", \".join(c[day][\"BDL\"])\n",
    "                axis_label = textwrap.fill(labels, width=30)\n",
    "                axar[n, i].set_ylabel(\"Colony forming units per 100mL\", fontdict={'fontsize': 7, 'fontweight': 'medium'},\n",
    "                                      labelpad=1,)\n",
    "                axar[n, i].set_xlabel('BDL : '+ str(axis_label))\n",
    "                axar[n, i].xaxis.set_major_locator(plt.NullLocator())\n",
    "                axar[n, i].xaxis.set_major_formatter(plt.NullFormatter())\n",
    "            else:\n",
    "                axar[n, i].bar(ind, c[day][\"Averages\"], width, color=colss)\n",
    "                axar[n, i].set_ylim(0, max(c[day][\"Averages\"]) + 50)\n",
    "                axar[n, i].set_title(b + ': ' +  day + ', samples: ' + str(c[day]['Number of Samples']) + ', ' +\n",
    "                                 str(c[day]['Qty plated'])+ 'mL', fontdict={'fontsize': 10, 'fontweight': 'medium'})\n",
    "                labels =  \", \".join(c[day][\"BDL\"])\n",
    "                axis_label = textwrap.fill(labels, width=30)\n",
    "                axar[n, i].set_xlabel('BDL : '+ str(axis_label))\n",
    "                axar[n, i].xaxis.set_major_locator(plt.NullLocator())\n",
    "                axar[n, i].xaxis.set_major_formatter(plt.NullFormatter())\n",
    "    \n",
    "    plt.annotate(title, xy=(0.5, .98),  xycoords=\"figure fraction\",                 \n",
    "                 size=18, ha='center', va=\"top\")\n",
    "    plt.annotate(subtitle, xy=(0.5, .96),  xycoords=\"figure fraction\",                 \n",
    "                 size=12, ha='center', va=\"top\")\n",
    "   \n",
    "    def makePatches(colors, legendColors):\n",
    "        a = []\n",
    "        for i,b in enumerate(colors):\n",
    "            c = mpatches.Patch(color=colors[i], label=legendColors[i])\n",
    "            a.append(c)\n",
    "        return a\n",
    "        \n",
    "    plt.legend(handles=makePatches(colors, legendColors),bbox_to_anchor=(0.11,.87, .8, .08),\n",
    "               mode=\"expand\",loc=\"center\", ncol=len(colors), bbox_transform=f.transFigure )\n",
    "    plt.savefig(svgtitle)\n",
    "    plt.show()\n",
    "    plt.close()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Define the variables for the charting function\n",
    "\n",
    "1. Correlate species to color\n",
    "2. Identify the colors you are going to use\n",
    "3. Make a title\n",
    "4. Make a subtitle"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [],
   "source": [
    "speciesColor16 = ['E.coli','Other coliforms', 'Salmonella','Aeromonas','other',]\n",
    "colors16 = ['darkblue', 'dodgerblue', 'turquoise', 'pink', 'lightslategray']\n",
    "title16 = \"Average 24 hour colony counts June-August 2016: ECA Check Plus\"\n",
    "subTitle16 = \"MRD = Montreux marché, SVT = Sauvetage, VNX = Vernex, BDL = Below detectable limit\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "plot_ez_totals16(theOutput16, 5, \"data/charts/2016_BARCHART_ARRAY.svg\", \"new2016.png\", title16, subTitle16, colors16, speciesColor16)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Export to JSON"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [],
   "source": [
    "def make_json(a, b):\n",
    "    with open(a, 'wb') as f:\n",
    "        f.write(json.dumps(b).encode('utf-8'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [],
   "source": [
    "make_json(\"data/JSON/APP_OUTPUT/barchartsArrays2017.json\", theOutput)\n",
    "make_json(\"data/JSON/APP_OUTPUT/barchartsArrays2016.json\",theOutput16)"
   ]
  }
 ],
 "metadata": {
  "hide_input": false,
  "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": "3.7.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}