{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "N2IUDwhPvDD2"
   },
   "source": [
    "# 2. Prepare input data of Yoshii River\n",
    "*Written by Men Vuthy, 2022*"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Import modules**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "executionInfo": {
     "elapsed": 2332,
     "status": "ok",
     "timestamp": 1647337947212,
     "user": {
      "displayName": "MEN Vuthy",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14Ggz_fgR6NAqa_UnWAIIDjX9pd1OkdeBn_u164cC=s64",
      "userId": "02183312642253897909"
     },
     "user_tz": -540
    },
    "id": "GUtVEQQsYOqG"
   },
   "outputs": [],
   "source": [
    "import os\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "import rasterio\n",
    "import geopandas as gpd"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "4kzODWJeteuT"
   },
   "source": [
    "**Label**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "executionInfo": {
     "elapsed": 533,
     "status": "ok",
     "timestamp": 1647337947742,
     "user": {
      "displayName": "MEN Vuthy",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14Ggz_fgR6NAqa_UnWAIIDjX9pd1OkdeBn_u164cC=s64",
      "userId": "02183312642253897909"
     },
     "user_tz": -540
    },
    "id": "0eRWOvq2YlB1"
   },
   "outputs": [],
   "source": [
    "yoshii_class_img = rasterio.open('data/yoshii_river/out_img/class/yoshii_class.tiff')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "executionInfo": {
     "elapsed": 3,
     "status": "ok",
     "timestamp": 1647337947742,
     "user": {
      "displayName": "MEN Vuthy",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14Ggz_fgR6NAqa_UnWAIIDjX9pd1OkdeBn_u164cC=s64",
      "userId": "02183312642253897909"
     },
     "user_tz": -540
    },
    "id": "8WuGITVVfVpR"
   },
   "outputs": [],
   "source": [
    "yoshii_class = yoshii_class_img.read(1).reshape(-1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "executionInfo": {
     "elapsed": 3,
     "status": "ok",
     "timestamp": 1647337947742,
     "user": {
      "displayName": "MEN Vuthy",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14Ggz_fgR6NAqa_UnWAIIDjX9pd1OkdeBn_u164cC=s64",
      "userId": "02183312642253897909"
     },
     "user_tz": -540
    },
    "id": "34uDVmPonYvn"
   },
   "outputs": [],
   "source": [
    "yoshii_label = pd.DataFrame({'label':yoshii_class})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "efyxwD0c8I8z"
   },
   "source": [
    "**Dataframe of Season 1**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "executionInfo": {
     "elapsed": 2,
     "status": "ok",
     "timestamp": 1647337947742,
     "user": {
      "displayName": "MEN Vuthy",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14Ggz_fgR6NAqa_UnWAIIDjX9pd1OkdeBn_u164cC=s64",
      "userId": "02183312642253897909"
     },
     "user_tz": -540
    },
    "id": "dxzPDKy_fvWT"
   },
   "outputs": [],
   "source": [
    "season1_rgbn = rasterio.open('data/yoshii_river/out_img/rgbn/yoshii_20190409_rgbn.tiff')\n",
    "season1_ndvi = rasterio.open('data/yoshii_river/out_img/ndvi/yoshii_20190409_ndvi.tiff')\n",
    "season1_ndwi = rasterio.open('data/yoshii_river/out_img/ndwi/yoshii_20190409_ndwi.tiff')\n",
    "season1_bsi = rasterio.open('data/yoshii_river/out_img/bsi/yoshii_20190409_bsi.tiff')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "executionInfo": {
     "elapsed": 2941,
     "status": "ok",
     "timestamp": 1647337950681,
     "user": {
      "displayName": "MEN Vuthy",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14Ggz_fgR6NAqa_UnWAIIDjX9pd1OkdeBn_u164cC=s64",
      "userId": "02183312642253897909"
     },
     "user_tz": -540
    },
    "id": "A9aRPiqjrBjC"
   },
   "outputs": [],
   "source": [
    "blue_1 = season1_rgbn.read(1).reshape(-1)\n",
    "green_1 = season1_rgbn.read(2).reshape(-1)\n",
    "red_1 = season1_rgbn.read(3).reshape(-1)\n",
    "nir_1 = season1_rgbn.read(4).reshape(-1)\n",
    "\n",
    "ndvi_1 = season1_ndvi.read(1).reshape(-1)\n",
    "ndwi_1 = season1_ndwi.read(1).reshape(-1)\n",
    "bsi_1 = season1_bsi.read(1).reshape(-1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "executionInfo": {
     "elapsed": 424,
     "status": "ok",
     "timestamp": 1647337951100,
     "user": {
      "displayName": "MEN Vuthy",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14Ggz_fgR6NAqa_UnWAIIDjX9pd1OkdeBn_u164cC=s64",
      "userId": "02183312642253897909"
     },
     "user_tz": -540
    },
    "id": "cKMuZhn8rJgR"
   },
   "outputs": [],
   "source": [
    "Season_1 = pd.DataFrame({'B1':blue_1, 'G1':green_1, 'R1':red_1, 'NIR1':nir_1, 'NDVI1':ndvi_1, 'NDWI1':ndwi_1, 'BSI1':bsi_1})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 423
    },
    "executionInfo": {
     "elapsed": 7,
     "status": "ok",
     "timestamp": 1647337951100,
     "user": {
      "displayName": "MEN Vuthy",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14Ggz_fgR6NAqa_UnWAIIDjX9pd1OkdeBn_u164cC=s64",
      "userId": "02183312642253897909"
     },
     "user_tz": -540
    },
    "id": "y8dt2IOysWrh",
    "outputId": "e4a20afc-68da-492d-9370-c313246520c8"
   },
   "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>B1</th>\n",
       "      <th>G1</th>\n",
       "      <th>R1</th>\n",
       "      <th>NIR1</th>\n",
       "      <th>NDVI1</th>\n",
       "      <th>NDWI1</th>\n",
       "      <th>BSI1</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38283940</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38283941</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38283942</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38283943</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38283944</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>38283945 rows × 7 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          B1  G1  R1  NIR1  NDVI1  NDWI1  BSI1\n",
       "0          0   0   0     0    NaN    NaN   NaN\n",
       "1          0   0   0     0    NaN    NaN   NaN\n",
       "2          0   0   0     0    NaN    NaN   NaN\n",
       "3          0   0   0     0    NaN    NaN   NaN\n",
       "4          0   0   0     0    NaN    NaN   NaN\n",
       "...       ..  ..  ..   ...    ...    ...   ...\n",
       "38283940   0   0   0     0    NaN    NaN   NaN\n",
       "38283941   0   0   0     0    NaN    NaN   NaN\n",
       "38283942   0   0   0     0    NaN    NaN   NaN\n",
       "38283943   0   0   0     0    NaN    NaN   NaN\n",
       "38283944   0   0   0     0    NaN    NaN   NaN\n",
       "\n",
       "[38283945 rows x 7 columns]"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Season_1"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "5tIbL4BkpJa2"
   },
   "source": [
    "**Dataframe of Season 2**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "executionInfo": {
     "elapsed": 271,
     "status": "ok",
     "timestamp": 1647337951367,
     "user": {
      "displayName": "MEN Vuthy",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14Ggz_fgR6NAqa_UnWAIIDjX9pd1OkdeBn_u164cC=s64",
      "userId": "02183312642253897909"
     },
     "user_tz": -540
    },
    "id": "yYKDT-cLpJJr"
   },
   "outputs": [],
   "source": [
    "season2_rgbn = rasterio.open('data/yoshii_river/out_img/rgbn/yoshii_20190805_rgbn.tiff')\n",
    "season2_ndvi = rasterio.open('data/yoshii_river/out_img/ndvi/yoshii_20190805_ndvi.tiff')\n",
    "season2_ndwi = rasterio.open('data/yoshii_river/out_img/ndwi/yoshii_20190805_ndwi.tiff')\n",
    "season2_bsi = rasterio.open('data/yoshii_river/out_img/bsi/yoshii_20190805_bsi.tiff')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "executionInfo": {
     "elapsed": 3676,
     "status": "ok",
     "timestamp": 1647337955040,
     "user": {
      "displayName": "MEN Vuthy",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14Ggz_fgR6NAqa_UnWAIIDjX9pd1OkdeBn_u164cC=s64",
      "userId": "02183312642253897909"
     },
     "user_tz": -540
    },
    "id": "lLq3mThwpZnI"
   },
   "outputs": [],
   "source": [
    "blue_2 = season2_rgbn.read(1).reshape(-1)\n",
    "green_2 = season2_rgbn.read(2).reshape(-1)\n",
    "red_2 = season2_rgbn.read(3).reshape(-1)\n",
    "nir_2 = season2_rgbn.read(4).reshape(-1)\n",
    "\n",
    "ndvi_2 = season2_ndvi.read(1).reshape(-1)\n",
    "ndwi_2 = season2_ndwi.read(1).reshape(-1)\n",
    "bsi_2 = season2_bsi.read(1).reshape(-1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "executionInfo": {
     "elapsed": 380,
     "status": "ok",
     "timestamp": 1647337955415,
     "user": {
      "displayName": "MEN Vuthy",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14Ggz_fgR6NAqa_UnWAIIDjX9pd1OkdeBn_u164cC=s64",
      "userId": "02183312642253897909"
     },
     "user_tz": -540
    },
    "id": "RAqkedgSpkXn"
   },
   "outputs": [],
   "source": [
    "Season_2 = pd.DataFrame({'B2':blue_2, 'G2':green_2, 'R2':red_2, 'NIR2':nir_2, 'NDVI2':ndvi_2, 'NDWI2':ndwi_2, 'BSI2':bsi_2})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 423
    },
    "executionInfo": {
     "elapsed": 5,
     "status": "ok",
     "timestamp": 1647337955416,
     "user": {
      "displayName": "MEN Vuthy",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14Ggz_fgR6NAqa_UnWAIIDjX9pd1OkdeBn_u164cC=s64",
      "userId": "02183312642253897909"
     },
     "user_tz": -540
    },
    "id": "j2Zp2QI2pru4",
    "outputId": "fab743f4-b152-4359-88bc-410cace9891d"
   },
   "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>B2</th>\n",
       "      <th>G2</th>\n",
       "      <th>R2</th>\n",
       "      <th>NIR2</th>\n",
       "      <th>NDVI2</th>\n",
       "      <th>NDWI2</th>\n",
       "      <th>BSI2</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38283940</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38283941</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38283942</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38283943</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38283944</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>38283945 rows × 7 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          B2  G2  R2  NIR2  NDVI2  NDWI2  BSI2\n",
       "0          0   0   0     0    NaN    NaN   NaN\n",
       "1          0   0   0     0    NaN    NaN   NaN\n",
       "2          0   0   0     0    NaN    NaN   NaN\n",
       "3          0   0   0     0    NaN    NaN   NaN\n",
       "4          0   0   0     0    NaN    NaN   NaN\n",
       "...       ..  ..  ..   ...    ...    ...   ...\n",
       "38283940   0   0   0     0    NaN    NaN   NaN\n",
       "38283941   0   0   0     0    NaN    NaN   NaN\n",
       "38283942   0   0   0     0    NaN    NaN   NaN\n",
       "38283943   0   0   0     0    NaN    NaN   NaN\n",
       "38283944   0   0   0     0    NaN    NaN   NaN\n",
       "\n",
       "[38283945 rows x 7 columns]"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Season_2"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "k_HvDVLoqCGI"
   },
   "source": [
    "**Dataframe of Season 3**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "executionInfo": {
     "elapsed": 4,
     "status": "ok",
     "timestamp": 1647337955416,
     "user": {
      "displayName": "MEN Vuthy",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14Ggz_fgR6NAqa_UnWAIIDjX9pd1OkdeBn_u164cC=s64",
      "userId": "02183312642253897909"
     },
     "user_tz": -540
    },
    "id": "_x0caSaXptHI"
   },
   "outputs": [],
   "source": [
    "season3_rgbn = rasterio.open('data/yoshii_river/out_img/rgbn/yoshii_20191007_rgbn.tiff')\n",
    "season3_ndvi = rasterio.open('data/yoshii_river/out_img/ndvi/yoshii_20191007_ndvi.tiff')\n",
    "season3_ndwi = rasterio.open('data/yoshii_river/out_img/ndwi/yoshii_20191007_ndwi.tiff')\n",
    "season3_bsi = rasterio.open('data/yoshii_river/out_img/bsi/yoshii_20191007_bsi.tiff')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "executionInfo": {
     "elapsed": 4134,
     "status": "ok",
     "timestamp": 1647337959546,
     "user": {
      "displayName": "MEN Vuthy",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14Ggz_fgR6NAqa_UnWAIIDjX9pd1OkdeBn_u164cC=s64",
      "userId": "02183312642253897909"
     },
     "user_tz": -540
    },
    "id": "7rdZVc-Mqige"
   },
   "outputs": [],
   "source": [
    "blue_3 = season3_rgbn.read(1).reshape(-1)\n",
    "green_3 = season3_rgbn.read(2).reshape(-1)\n",
    "red_3 = season3_rgbn.read(3).reshape(-1)\n",
    "nir_3 = season3_rgbn.read(4).reshape(-1)\n",
    "\n",
    "ndvi_3 = season3_ndvi.read(1).reshape(-1)\n",
    "ndwi_3 = season3_ndwi.read(1).reshape(-1)\n",
    "bsi_3 = season3_bsi.read(1).reshape(-1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "executionInfo": {
     "elapsed": 1324,
     "status": "ok",
     "timestamp": 1647337960860,
     "user": {
      "displayName": "MEN Vuthy",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14Ggz_fgR6NAqa_UnWAIIDjX9pd1OkdeBn_u164cC=s64",
      "userId": "02183312642253897909"
     },
     "user_tz": -540
    },
    "id": "cUdE7Gq3qpVX"
   },
   "outputs": [],
   "source": [
    "Season_3 = pd.DataFrame({'B3':blue_3, 'G3':green_3, 'R3':red_3, 'NIR3':nir_3, 'NDVI3':ndvi_3, 'NDWI3':ndwi_3, 'BSI3':bsi_3})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 423
    },
    "executionInfo": {
     "elapsed": 5,
     "status": "ok",
     "timestamp": 1647337960860,
     "user": {
      "displayName": "MEN Vuthy",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14Ggz_fgR6NAqa_UnWAIIDjX9pd1OkdeBn_u164cC=s64",
      "userId": "02183312642253897909"
     },
     "user_tz": -540
    },
    "id": "ETLVsonaq0qo",
    "outputId": "a3ca1325-c6f4-4a43-8278-9ce6c2841ac7"
   },
   "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>B3</th>\n",
       "      <th>G3</th>\n",
       "      <th>R3</th>\n",
       "      <th>NIR3</th>\n",
       "      <th>NDVI3</th>\n",
       "      <th>NDWI3</th>\n",
       "      <th>BSI3</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38283940</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38283941</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38283942</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38283943</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38283944</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>38283945 rows × 7 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          B3  G3  R3  NIR3  NDVI3  NDWI3  BSI3\n",
       "0          0   0   0     0    NaN    NaN   NaN\n",
       "1          0   0   0     0    NaN    NaN   NaN\n",
       "2          0   0   0     0    NaN    NaN   NaN\n",
       "3          0   0   0     0    NaN    NaN   NaN\n",
       "4          0   0   0     0    NaN    NaN   NaN\n",
       "...       ..  ..  ..   ...    ...    ...   ...\n",
       "38283940   0   0   0     0    NaN    NaN   NaN\n",
       "38283941   0   0   0     0    NaN    NaN   NaN\n",
       "38283942   0   0   0     0    NaN    NaN   NaN\n",
       "38283943   0   0   0     0    NaN    NaN   NaN\n",
       "38283944   0   0   0     0    NaN    NaN   NaN\n",
       "\n",
       "[38283945 rows x 7 columns]"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Season_3"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "S2-5njKqq-Ou"
   },
   "source": [
    "**Dataframe of Season 4**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "executionInfo": {
     "elapsed": 5,
     "status": "ok",
     "timestamp": 1647337960861,
     "user": {
      "displayName": "MEN Vuthy",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14Ggz_fgR6NAqa_UnWAIIDjX9pd1OkdeBn_u164cC=s64",
      "userId": "02183312642253897909"
     },
     "user_tz": -540
    },
    "id": "RiTAGhkjq3T4"
   },
   "outputs": [],
   "source": [
    "season4_rgbn = rasterio.open('data/yoshii_river/out_img/rgbn/yoshii_20200104_rgbn.tiff')\n",
    "season4_ndvi = rasterio.open('data/yoshii_river/out_img/ndvi/yoshii_20200104_ndvi.tiff')\n",
    "season4_ndwi = rasterio.open('data/yoshii_river/out_img/ndwi/yoshii_20200104_ndwi.tiff')\n",
    "season4_bsi = rasterio.open('data/yoshii_river/out_img/bsi/yoshii_20200104_bsi.tiff')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "executionInfo": {
     "elapsed": 5373,
     "status": "ok",
     "timestamp": 1647337966229,
     "user": {
      "displayName": "MEN Vuthy",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14Ggz_fgR6NAqa_UnWAIIDjX9pd1OkdeBn_u164cC=s64",
      "userId": "02183312642253897909"
     },
     "user_tz": -540
    },
    "id": "7Tb5TVotrQc4"
   },
   "outputs": [],
   "source": [
    "blue_4 = season4_rgbn.read(1).reshape(-1)\n",
    "green_4 = season4_rgbn.read(2).reshape(-1)\n",
    "red_4 = season4_rgbn.read(3).reshape(-1)\n",
    "nir_4 = season4_rgbn.read(4).reshape(-1)\n",
    "\n",
    "ndvi_4 = season4_ndvi.read(1).reshape(-1)\n",
    "ndwi_4 = season4_ndwi.read(1).reshape(-1)\n",
    "bsi_4 = season4_bsi.read(1).reshape(-1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "executionInfo": {
     "elapsed": 1067,
     "status": "ok",
     "timestamp": 1647337967285,
     "user": {
      "displayName": "MEN Vuthy",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14Ggz_fgR6NAqa_UnWAIIDjX9pd1OkdeBn_u164cC=s64",
      "userId": "02183312642253897909"
     },
     "user_tz": -540
    },
    "id": "Ly1uhXo4rXX3"
   },
   "outputs": [],
   "source": [
    "Season_4 = pd.DataFrame({'B4':blue_4, 'G4':green_4, 'R4':red_4, 'NIR4':nir_4, 'NDVI4':ndvi_4, 'NDWI4':ndwi_4, 'BSI4':bsi_4})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 423
    },
    "executionInfo": {
     "elapsed": 380,
     "status": "ok",
     "timestamp": 1647337967663,
     "user": {
      "displayName": "MEN Vuthy",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14Ggz_fgR6NAqa_UnWAIIDjX9pd1OkdeBn_u164cC=s64",
      "userId": "02183312642253897909"
     },
     "user_tz": -540
    },
    "id": "p_ivo4ZordUW",
    "outputId": "2f625008-73c0-4486-9aef-393a865a0d5d"
   },
   "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>B4</th>\n",
       "      <th>G4</th>\n",
       "      <th>R4</th>\n",
       "      <th>NIR4</th>\n",
       "      <th>NDVI4</th>\n",
       "      <th>NDWI4</th>\n",
       "      <th>BSI4</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38283940</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38283941</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38283942</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38283943</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38283944</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>38283945 rows × 7 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          B4  G4  R4  NIR4  NDVI4  NDWI4  BSI4\n",
       "0          0   0   0     0    NaN    NaN   NaN\n",
       "1          0   0   0     0    NaN    NaN   NaN\n",
       "2          0   0   0     0    NaN    NaN   NaN\n",
       "3          0   0   0     0    NaN    NaN   NaN\n",
       "4          0   0   0     0    NaN    NaN   NaN\n",
       "...       ..  ..  ..   ...    ...    ...   ...\n",
       "38283940   0   0   0     0    NaN    NaN   NaN\n",
       "38283941   0   0   0     0    NaN    NaN   NaN\n",
       "38283942   0   0   0     0    NaN    NaN   NaN\n",
       "38283943   0   0   0     0    NaN    NaN   NaN\n",
       "38283944   0   0   0     0    NaN    NaN   NaN\n",
       "\n",
       "[38283945 rows x 7 columns]"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Season_4"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "LhqoBmmMvczS"
   },
   "source": [
    "**Features and Label**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "executionInfo": {
     "elapsed": 4323,
     "status": "ok",
     "timestamp": 1647337971981,
     "user": {
      "displayName": "MEN Vuthy",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14Ggz_fgR6NAqa_UnWAIIDjX9pd1OkdeBn_u164cC=s64",
      "userId": "02183312642253897909"
     },
     "user_tz": -540
    },
    "id": "-MMyjJRkrfjX"
   },
   "outputs": [],
   "source": [
    "# concat dataframe of all seasons and label into one dataframe\n",
    "Yoshii_dataframe = pd.concat([Season_1, Season_2, Season_3, Season_4, yoshii_label], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "id": "z9jaUKUfv10G"
   },
   "outputs": [],
   "source": [
    "# Remove all rows where NDVI1 is NaN\n",
    "Yoshii_classified = Yoshii_dataframe.dropna(subset=['NDVI1'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "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>B1</th>\n",
       "      <th>G1</th>\n",
       "      <th>R1</th>\n",
       "      <th>NIR1</th>\n",
       "      <th>NDVI1</th>\n",
       "      <th>NDWI1</th>\n",
       "      <th>BSI1</th>\n",
       "      <th>B2</th>\n",
       "      <th>G2</th>\n",
       "      <th>R2</th>\n",
       "      <th>...</th>\n",
       "      <th>NDWI3</th>\n",
       "      <th>BSI3</th>\n",
       "      <th>B4</th>\n",
       "      <th>G4</th>\n",
       "      <th>R4</th>\n",
       "      <th>NIR4</th>\n",
       "      <th>NDVI4</th>\n",
       "      <th>NDWI4</th>\n",
       "      <th>BSI4</th>\n",
       "      <th>label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>4371353</th>\n",
       "      <td>558</td>\n",
       "      <td>601</td>\n",
       "      <td>513</td>\n",
       "      <td>889</td>\n",
       "      <td>-0.023003</td>\n",
       "      <td>0.081056</td>\n",
       "      <td>0.279785</td>\n",
       "      <td>324</td>\n",
       "      <td>421</td>\n",
       "      <td>385</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.054566</td>\n",
       "      <td>0.235740</td>\n",
       "      <td>349</td>\n",
       "      <td>489</td>\n",
       "      <td>457</td>\n",
       "      <td>686</td>\n",
       "      <td>0.135323</td>\n",
       "      <td>-0.030485</td>\n",
       "      <td>0.222090</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4371354</th>\n",
       "      <td>558</td>\n",
       "      <td>601</td>\n",
       "      <td>513</td>\n",
       "      <td>889</td>\n",
       "      <td>-0.023003</td>\n",
       "      <td>0.081056</td>\n",
       "      <td>0.279785</td>\n",
       "      <td>324</td>\n",
       "      <td>421</td>\n",
       "      <td>385</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.054566</td>\n",
       "      <td>0.235740</td>\n",
       "      <td>349</td>\n",
       "      <td>489</td>\n",
       "      <td>457</td>\n",
       "      <td>686</td>\n",
       "      <td>0.135323</td>\n",
       "      <td>-0.030485</td>\n",
       "      <td>0.222090</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4371355</th>\n",
       "      <td>614</td>\n",
       "      <td>639</td>\n",
       "      <td>533</td>\n",
       "      <td>1012</td>\n",
       "      <td>-0.006031</td>\n",
       "      <td>0.047061</td>\n",
       "      <td>0.282591</td>\n",
       "      <td>334</td>\n",
       "      <td>440</td>\n",
       "      <td>420</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.118004</td>\n",
       "      <td>0.259052</td>\n",
       "      <td>387</td>\n",
       "      <td>540</td>\n",
       "      <td>571</td>\n",
       "      <td>843</td>\n",
       "      <td>0.127158</td>\n",
       "      <td>-0.083739</td>\n",
       "      <td>0.253888</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4371356</th>\n",
       "      <td>705</td>\n",
       "      <td>715</td>\n",
       "      <td>606</td>\n",
       "      <td>1203</td>\n",
       "      <td>0.011312</td>\n",
       "      <td>0.016838</td>\n",
       "      <td>0.292253</td>\n",
       "      <td>391</td>\n",
       "      <td>507</td>\n",
       "      <td>509</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.222598</td>\n",
       "      <td>0.262350</td>\n",
       "      <td>458</td>\n",
       "      <td>627</td>\n",
       "      <td>705</td>\n",
       "      <td>1027</td>\n",
       "      <td>0.120572</td>\n",
       "      <td>-0.107544</td>\n",
       "      <td>0.273603</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4371357</th>\n",
       "      <td>816</td>\n",
       "      <td>794</td>\n",
       "      <td>681</td>\n",
       "      <td>1472</td>\n",
       "      <td>0.039086</td>\n",
       "      <td>-0.031651</td>\n",
       "      <td>0.304678</td>\n",
       "      <td>501</td>\n",
       "      <td>614</td>\n",
       "      <td>648</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.257952</td>\n",
       "      <td>0.264777</td>\n",
       "      <td>557</td>\n",
       "      <td>733</td>\n",
       "      <td>838</td>\n",
       "      <td>1254</td>\n",
       "      <td>0.133792</td>\n",
       "      <td>-0.128988</td>\n",
       "      <td>0.285978</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34076326</th>\n",
       "      <td>896</td>\n",
       "      <td>883</td>\n",
       "      <td>753</td>\n",
       "      <td>2833</td>\n",
       "      <td>0.309235</td>\n",
       "      <td>-0.296701</td>\n",
       "      <td>0.300410</td>\n",
       "      <td>473</td>\n",
       "      <td>619</td>\n",
       "      <td>575</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.482353</td>\n",
       "      <td>0.253913</td>\n",
       "      <td>492</td>\n",
       "      <td>664</td>\n",
       "      <td>795</td>\n",
       "      <td>1464</td>\n",
       "      <td>0.233941</td>\n",
       "      <td>-0.251074</td>\n",
       "      <td>0.292699</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34076327</th>\n",
       "      <td>919</td>\n",
       "      <td>904</td>\n",
       "      <td>775</td>\n",
       "      <td>2700</td>\n",
       "      <td>0.275665</td>\n",
       "      <td>-0.263721</td>\n",
       "      <td>0.301996</td>\n",
       "      <td>484</td>\n",
       "      <td>628</td>\n",
       "      <td>582</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.464620</td>\n",
       "      <td>0.253741</td>\n",
       "      <td>489</td>\n",
       "      <td>667</td>\n",
       "      <td>818</td>\n",
       "      <td>1428</td>\n",
       "      <td>0.208541</td>\n",
       "      <td>-0.237249</td>\n",
       "      <td>0.296889</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34076328</th>\n",
       "      <td>949</td>\n",
       "      <td>929</td>\n",
       "      <td>803</td>\n",
       "      <td>2563</td>\n",
       "      <td>0.236333</td>\n",
       "      <td>-0.226434</td>\n",
       "      <td>0.304925</td>\n",
       "      <td>523</td>\n",
       "      <td>659</td>\n",
       "      <td>616</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.424649</td>\n",
       "      <td>0.263849</td>\n",
       "      <td>519</td>\n",
       "      <td>704</td>\n",
       "      <td>875</td>\n",
       "      <td>1397</td>\n",
       "      <td>0.165456</td>\n",
       "      <td>-0.201109</td>\n",
       "      <td>0.301471</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34076329</th>\n",
       "      <td>964</td>\n",
       "      <td>951</td>\n",
       "      <td>827</td>\n",
       "      <td>2445</td>\n",
       "      <td>0.206467</td>\n",
       "      <td>-0.192717</td>\n",
       "      <td>0.304421</td>\n",
       "      <td>559</td>\n",
       "      <td>695</td>\n",
       "      <td>657</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.351897</td>\n",
       "      <td>0.268711</td>\n",
       "      <td>507</td>\n",
       "      <td>736</td>\n",
       "      <td>859</td>\n",
       "      <td>1422</td>\n",
       "      <td>0.183002</td>\n",
       "      <td>-0.188259</td>\n",
       "      <td>0.271648</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34076330</th>\n",
       "      <td>1010</td>\n",
       "      <td>989</td>\n",
       "      <td>869</td>\n",
       "      <td>2362</td>\n",
       "      <td>0.167315</td>\n",
       "      <td>-0.156990</td>\n",
       "      <td>0.308427</td>\n",
       "      <td>615</td>\n",
       "      <td>750</td>\n",
       "      <td>733</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.318919</td>\n",
       "      <td>0.269634</td>\n",
       "      <td>573</td>\n",
       "      <td>791</td>\n",
       "      <td>897</td>\n",
       "      <td>1501</td>\n",
       "      <td>0.188206</td>\n",
       "      <td>-0.179563</td>\n",
       "      <td>0.274091</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1152991 rows × 29 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "            B1   G1   R1  NIR1     NDVI1     NDWI1      BSI1   B2   G2   R2  \\\n",
       "4371353    558  601  513   889 -0.023003  0.081056  0.279785  324  421  385   \n",
       "4371354    558  601  513   889 -0.023003  0.081056  0.279785  324  421  385   \n",
       "4371355    614  639  533  1012 -0.006031  0.047061  0.282591  334  440  420   \n",
       "4371356    705  715  606  1203  0.011312  0.016838  0.292253  391  507  509   \n",
       "4371357    816  794  681  1472  0.039086 -0.031651  0.304678  501  614  648   \n",
       "...        ...  ...  ...   ...       ...       ...       ...  ...  ...  ...   \n",
       "34076326   896  883  753  2833  0.309235 -0.296701  0.300410  473  619  575   \n",
       "34076327   919  904  775  2700  0.275665 -0.263721  0.301996  484  628  582   \n",
       "34076328   949  929  803  2563  0.236333 -0.226434  0.304925  523  659  616   \n",
       "34076329   964  951  827  2445  0.206467 -0.192717  0.304421  559  695  657   \n",
       "34076330  1010  989  869  2362  0.167315 -0.156990  0.308427  615  750  733   \n",
       "\n",
       "          ...     NDWI3      BSI3   B4   G4   R4  NIR4     NDVI4     NDWI4  \\\n",
       "4371353   ... -0.054566  0.235740  349  489  457   686  0.135323 -0.030485   \n",
       "4371354   ... -0.054566  0.235740  349  489  457   686  0.135323 -0.030485   \n",
       "4371355   ... -0.118004  0.259052  387  540  571   843  0.127158 -0.083739   \n",
       "4371356   ... -0.222598  0.262350  458  627  705  1027  0.120572 -0.107544   \n",
       "4371357   ... -0.257952  0.264777  557  733  838  1254  0.133792 -0.128988   \n",
       "...       ...       ...       ...  ...  ...  ...   ...       ...       ...   \n",
       "34076326  ... -0.482353  0.253913  492  664  795  1464  0.233941 -0.251074   \n",
       "34076327  ... -0.464620  0.253741  489  667  818  1428  0.208541 -0.237249   \n",
       "34076328  ... -0.424649  0.263849  519  704  875  1397  0.165456 -0.201109   \n",
       "34076329  ... -0.351897  0.268711  507  736  859  1422  0.183002 -0.188259   \n",
       "34076330  ... -0.318919  0.269634  573  791  897  1501  0.188206 -0.179563   \n",
       "\n",
       "              BSI4  label  \n",
       "4371353   0.222090      7  \n",
       "4371354   0.222090      7  \n",
       "4371355   0.253888      7  \n",
       "4371356   0.273603      1  \n",
       "4371357   0.285978      1  \n",
       "...            ...    ...  \n",
       "34076326  0.292699      1  \n",
       "34076327  0.296889      1  \n",
       "34076328  0.301471      1  \n",
       "34076329  0.271648      1  \n",
       "34076330  0.274091      1  \n",
       "\n",
       "[1152991 rows x 29 columns]"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Yoshii_classified"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "id": "kfDWqViT1pRI"
   },
   "outputs": [],
   "source": [
    "# Data dir\n",
    "data_dir = \"data/yoshii_river/out_img/classified\"\n",
    "\n",
    "# Output dataframe\n",
    "out_df = os.path.join(data_dir, 'yoshii_classified.csv')\n",
    "\n",
    "# export dataframe to csv\n",
    "Yoshii_classified.to_csv(out_df, index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "id": "ut1eM8opzDWU"
   },
   "outputs": [],
   "source": [
    "# Data dir\n",
    "data_dir = \"data/yoshii_river/out_img/classified\"\n",
    "\n",
    "# Output dataframe\n",
    "out_df = os.path.join(data_dir, 'yoshii_classified_index.csv')\n",
    "\n",
    "# export dataframe to csv\n",
    "Yoshii_classified.to_csv(out_df)"
   ]
  }
 ],
 "metadata": {
  "colab": {
   "authorship_tag": "ABX9TyO3G0LDfhmDKTCllN3BGQnY",
   "name": "6. Input Data for Supervised-ML- Yoshii River.ipynb",
   "provenance": []
  },
  "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": "3.7.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}