Supervised_Classification_of_IMG_DATA_1.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Code adapted from https://github.com/patrickcgray/open-geo-tutorial"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 178,
   "metadata": {},
   "outputs": [],
   "source": [
    "from IPython.display import Audio, display\n",
    "from timeit import default_timer as timer\n",
    "start = timer()\n",
    "def alert():\n",
    "  global start\n",
    "  end= timer()\n",
    "  print(\"\\n\\nDuration in minutes : \" +  str((end - start)/60) + \" minutes.\" )\n",
    "  display(Audio(url='https://sound.peal.io/ps/audios/000/000/537/original/woo_vu_luvub_dub_dub.wav', autoplay=True))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Preparing Dataset"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Cutting images for faster process"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 179,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 0 to  10980 for this img_data\n",
    "#training\n",
    "xmin = 6500\n",
    "xmax = 7000\n",
    "ymin = 6600\n",
    "ymax = 7100\n",
    "#test\n",
    "xmin_test = 6000\n",
    "xmax_test = 6500\n",
    "ymin_test = 6800\n",
    "ymax_test = 7300"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 180,
   "metadata": {},
   "outputs": [],
   "source": [
    "import rasterio\n",
    "from rasterio.mask import mask\n",
    "import geopandas as gp\n",
    "import numpy as np\n",
    "import shapely\n",
    "from shapely import geometry\n",
    "from shapely.geometry import shape, Point, LineString, Polygon , mapping\n",
    "import matplotlib.pyplot as plt\n",
    "from rasterio.plot import show\n",
    "import pyproj\n",
    "from pyproj import CRS\n",
    "import fiona"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 181,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['./data/IMG_DATA_1\\\\T29SMD_20191025T113321_B02.jp2',\n",
       " './data/IMG_DATA_1\\\\T29SMD_20191025T113321_B03.jp2',\n",
       " './data/IMG_DATA_1\\\\T29SMD_20191025T113321_B04.jp2',\n",
       " './data/IMG_DATA_1\\\\T29SMD_20191025T113321_B08.jp2']"
      ]
     },
     "execution_count": 181,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import os # we need os to do some basic file operations\n",
    "import re\n",
    "regexp = re.compile(r'_B0[2348]')\n",
    "sentinal_20m = \"./data/IMG_DATA_1\"\n",
    "# find every file in the sentinal_fp directory\n",
    "sentinal_band_paths = [os.path.join(sentinal_20m, f) for f in os.listdir(sentinal_20m) if regexp.search(f) ]\n",
    "sentinal_band_paths.sort()\n",
    "sentinal_band_paths"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 182,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "./data/IMG_DATA_1\\T29SMD_20191025T113321_B02.jp2 ::: \n",
      "{'driver': 'JP2OpenJPEG', 'dtype': 'uint16', 'nodata': None, 'width': 10980, 'height': 10980, 'count': 1, 'crs': CRS.from_epsg(32629), 'transform': Affine(10.0, 0.0, 399960.0,\n",
      "       0.0, -10.0, 4400040.0)}\n",
      "./data/IMG_DATA_1\\T29SMD_20191025T113321_B03.jp2 ::: \n",
      "{'driver': 'JP2OpenJPEG', 'dtype': 'uint16', 'nodata': None, 'width': 10980, 'height': 10980, 'count': 1, 'crs': CRS.from_epsg(32629), 'transform': Affine(10.0, 0.0, 399960.0,\n",
      "       0.0, -10.0, 4400040.0)}\n",
      "./data/IMG_DATA_1\\T29SMD_20191025T113321_B04.jp2 ::: \n",
      "{'driver': 'JP2OpenJPEG', 'dtype': 'uint16', 'nodata': None, 'width': 10980, 'height': 10980, 'count': 1, 'crs': CRS.from_epsg(32629), 'transform': Affine(10.0, 0.0, 399960.0,\n",
      "       0.0, -10.0, 4400040.0)}\n",
      "./data/IMG_DATA_1\\T29SMD_20191025T113321_B08.jp2 ::: \n",
      "{'driver': 'JP2OpenJPEG', 'dtype': 'uint16', 'nodata': None, 'width': 10980, 'height': 10980, 'count': 1, 'crs': CRS.from_epsg(32629), 'transform': Affine(10.0, 0.0, 399960.0,\n",
      "       0.0, -10.0, 4400040.0)}\n"
     ]
    }
   ],
   "source": [
    "for i in sentinal_band_paths : \n",
    "        src = rasterio.open(i,driver='JP2OpenJPEG')\n",
    "        meta = src.meta\n",
    "        print( i + \" ::: \")\n",
    "        print(meta)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 183,
   "metadata": {},
   "outputs": [],
   "source": [
    "# create a products directory within the data dir which won't be uploaded to Github\n",
    "img_dir = './data/'\n",
    "\n",
    "# check to see if the dir it exists, if not, create it\n",
    "if not os.path.exists(img_dir):\n",
    "    os.makedirs(img_dir)\n",
    "\n",
    "# filepath for image we're writing out\n",
    "img_fp = img_dir + 'sentinel_bands_img_data_1.tif'"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Only run this the first time - its slow"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 184,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\"\\n# Just run this the first time \\nmeta = meta\\n# Update metadata to reflect the number of layers\\nmeta.update(\\n    count = len(sentinal_band_paths)\\n)\\n# Read each layer and write it to stack\\nwith rasterio.open(img_fp, 'w', **meta) as dst:\\n    for id, layer in enumerate(sentinal_band_paths, start=1):\\n        with rasterio.open(layer) as src1:\\n            dst.write_band(id, src1.read(1))\\n\""
      ]
     },
     "execution_count": 184,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\"\"\"\n",
    "# Just run this the first time \n",
    "meta = meta\n",
    "# Update metadata to reflect the number of layers\n",
    "meta.update(\n",
    "    count = len(sentinal_band_paths)\n",
    ")\n",
    "# Read each layer and write it to stack\n",
    "with rasterio.open(img_fp, 'w', **meta) as dst:\n",
    "    for id, layer in enumerate(sentinal_band_paths, start=1):\n",
    "        with rasterio.open(layer) as src1:\n",
    "            dst.write_band(id, src1.read(1))\n",
    "\"\"\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 185,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(10980, 10980)\n",
      "4\n"
     ]
    }
   ],
   "source": [
    "full_dataset = rasterio.open(img_fp)\n",
    "img_rows, img_cols = full_dataset.shape\n",
    "img_bands = full_dataset.count\n",
    "print(full_dataset.shape) # dimensions\n",
    "print(full_dataset.count) # bands"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 186,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Image filename: ./data/sentinel_bands_img_data_1.tif\n",
      "\n",
      "Number of bands in image: 4\n",
      "\n",
      "Image size is: 10980 rows x 10980 columns\n",
      "\n",
      "Raster description: (None, None, None, None)\n",
      "\n",
      "Raster driver: JP2OpenJPEG\n",
      "\n",
      "Image projection:\n",
      "EPSG:32629 \n",
      "\n",
      "Image geo-transform:\n",
      "| 10.00, 0.00, 399960.00|\n",
      "| 0.00,-10.00, 4400040.00|\n",
      "| 0.00, 0.00, 1.00|\n",
      "\n",
      "All raster metadata:\n",
      "{'driver': 'JP2OpenJPEG', 'dtype': 'uint16', 'nodata': None, 'width': 10980, 'height': 10980, 'count': 4, 'crs': CRS.from_epsg(32629), 'transform': Affine(10.0, 0.0, 399960.0,\n",
      "       0.0, -10.0, 4400040.0)}\n",
      "\n",
      "\n"
     ]
    }
   ],
   "source": [
    "dataset=full_dataset\n",
    "# what is the name of this image\n",
    "img_name = dataset.name\n",
    "print('Image filename: {n}\\n'.format(n=img_name))\n",
    "\n",
    "# How many bands does this image have?\n",
    "num_bands = dataset.count\n",
    "print('Number of bands in image: {n}\\n'.format(n=num_bands))\n",
    "\n",
    "# How many rows and columns?\n",
    "rows, cols = dataset.shape\n",
    "print('Image size is: {r} rows x {c} columns\\n'.format(r=rows, c=cols))\n",
    "\n",
    "# Does the raster have a description or metadata?\n",
    "desc = dataset.descriptions\n",
    "metadata = dataset.meta\n",
    "\n",
    "print('Raster description: {desc}\\n'.format(desc=desc))\n",
    "\n",
    "# What driver was used to open the raster?\n",
    "driver = dataset.driver\n",
    "print('Raster driver: {d}\\n'.format(d=driver))\n",
    "\n",
    "# What is the raster's projection?\n",
    "proj = dataset.crs\n",
    "if (not proj) :\n",
    "    proj = CRS.from_epsg(4326)\n",
    "print('Image projection:')\n",
    "print(proj, '\\n')\n",
    "\n",
    "# What is the raster's \"geo-transform\"\n",
    "gt = dataset.transform\n",
    "\n",
    "print('Image geo-transform:\\n{gt}\\n'.format(gt=gt))\n",
    "\n",
    "print('All raster metadata:')\n",
    "print(metadata)\n",
    "print('\\n')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 187,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x23001b35668>"
      ]
     },
     "execution_count": 187,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from rasterio.plot import adjust_band\n",
    "from rasterio.plot import reshape_as_raster, reshape_as_image\n",
    "\n",
    "\n",
    "cutted = dataset.read([3,2,1])[:,xmin: xmax,ymin : ymax]\n",
    "rgb_norm = adjust_band(cutted) # normalize bands to range between 1.0 to 0.0\n",
    "show(rgb_norm)\n",
    "\n",
    "cutted = dataset.read([3,2,1])[:,xmin_test: xmax_test,ymin_test : ymax_test]\n",
    "rgb_norm = adjust_band(cutted) # normalize bands to range between 1.0 to 0.0\n",
    "\n",
    "show(rgb_norm)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 188,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'\\nclipped_img = adjust_band(cutted)\\nprint(clipped_img.shape)\\nfig, ax = plt.subplots(figsize=(10,7))\\nshow(clipped_img[:,  : , : ], ax=ax, transform=full_dataset.transform) # add the transform arg to get it in lat long coords\\n'"
      ]
     },
     "execution_count": 188,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "from rasterio.plot import show\n",
    "\"\"\"\n",
    "clipped_img = adjust_band(cutted)\n",
    "print(clipped_img.shape)\n",
    "fig, ax = plt.subplots(figsize=(10,7))\n",
    "show(clipped_img[:,  : , : ], ax=ax, transform=full_dataset.transform) # add the transform arg to get it in lat long coords\n",
    "\"\"\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Indices"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 189,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "with rasterio.open(img_fp) as src:\n",
    "    green_band = src.read(2)[ xmin : xmax ,  ymin : ymax ] # band 3\n",
    "    red_band = src.read(3)[ xmin : xmax ,  ymin : ymax ]   # band 4\n",
    "    nir_band = src.read(4)[ xmin : xmax ,  ymin : ymax ]   # band 8\n",
    "    img = src.read()[:,  xmin : xmax ,  ymin : ymax ]\n",
    "    \n",
    "ndwi = (green_band.astype(float) - nir_band.astype(float)) / (green_band.astype(float) + nir_band.astype(float))\n",
    "ndvi = (nir_band.astype(float) - red_band.astype(float)) / (red_band.astype(float) + nir_band.astype(float))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 190,
   "metadata": {},
   "outputs": [],
   "source": [
    "def color_stretch(image, index):\n",
    "    colors = image[:, :, index].astype(np.float64)\n",
    "    for b in range(colors.shape[2]):\n",
    "        colors[:, :, b] = rasterio.plot.adjust_band(colors[:, :, b])\n",
    "    return colors"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 191,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'\\nfig, axs = plt.subplots(2,2,figsize=(15,7))\\n\\nreshaped_img =reshape_as_image( img )\\nimg_stretched = color_stretch(reshaped_img, [2, 1, 0])\\naxs[0,0].set_title(\"Stretch\")\\naxs[0,0].imshow(img_stretched)\\n\\naxs[0,1].set_title(\"True Color\")\\naxs[0,1].imshow(reshape_as_image(rgb_norm))\\n\\nnwdi_plot = axs[1,0].imshow(ndwi, cmap=\"RdYlBu\")\\naxs[1,0].set_title(\"NDWI\")\\nfig.colorbar(nwdi_plot, ax=axs[1,0])\\n\\nndvi_plot = axs[1,1].imshow(ndvi, cmap=\"RdYlGn\")\\naxs[1,1].set_title(\"NDVI\")\\nfig.colorbar(ndvi_plot, ax=axs[1,1])\\n\\nplt.show()\\n'"
      ]
     },
     "execution_count": 191,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\"\"\"\n",
    "fig, axs = plt.subplots(2,2,figsize=(15,7))\n",
    "\n",
    "reshaped_img =reshape_as_image( img )\n",
    "img_stretched = color_stretch(reshaped_img, [2, 1, 0])\n",
    "axs[0,0].set_title(\"Stretch\")\n",
    "axs[0,0].imshow(img_stretched)\n",
    "\n",
    "axs[0,1].set_title(\"True Color\")\n",
    "axs[0,1].imshow(reshape_as_image(rgb_norm))\n",
    "\n",
    "nwdi_plot = axs[1,0].imshow(ndwi, cmap=\"RdYlBu\")\n",
    "axs[1,0].set_title(\"NDWI\")\n",
    "fig.colorbar(nwdi_plot, ax=axs[1,0])\n",
    "\n",
    "ndvi_plot = axs[1,1].imshow(ndvi, cmap=\"RdYlGn\")\n",
    "axs[1,1].set_title(\"NDVI\")\n",
    "fig.colorbar(ndvi_plot, ax=axs[1,1])\n",
    "\n",
    "plt.show()\n",
    "\"\"\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 192,
   "metadata": {},
   "outputs": [],
   "source": [
    "src.close()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Intercepting label data with cutting image"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 193,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Open the dataset from the file\n",
    "shapefile = gp.read_file('./data/shapefiles/shapefile_compressed/shp_2012_IMG_DATA_1.shp')\n",
    "shapefile = shapefile.filter(['Legenda','geometry','CLC2012'])\n",
    "shapefile.crs = {'init': 'epsg:4326'}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 194,
   "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>Legenda</th>\n",
       "      <th>geometry</th>\n",
       "      <th>CLC2012</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Tecido urbano contínuo</td>\n",
       "      <td>POLYGON ((-9.27768 38.76588, -9.27746 38.76555...</td>\n",
       "      <td>111</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Tecido urbano contínuo</td>\n",
       "      <td>POLYGON ((-9.25850 38.76457, -9.25850 38.76158...</td>\n",
       "      <td>111</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Tecido urbano contínuo</td>\n",
       "      <td>POLYGON ((-9.21069 38.76698, -9.21064 38.76633...</td>\n",
       "      <td>111</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Tecido urbano contínuo</td>\n",
       "      <td>POLYGON ((-9.22487 38.76998, -9.22474 38.76907...</td>\n",
       "      <td>111</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Tecido urbano contínuo</td>\n",
       "      <td>POLYGON ((-9.15237 38.77167, -9.15397 38.77135...</td>\n",
       "      <td>111</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2606</th>\n",
       "      <td>Planos de água</td>\n",
       "      <td>POLYGON ((-9.30538 39.32667, -9.30438 39.32591...</td>\n",
       "      <td>512</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2607</th>\n",
       "      <td>Planos de água</td>\n",
       "      <td>POLYGON ((-9.12876 39.36121, -9.12819 39.36034...</td>\n",
       "      <td>512</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2608</th>\n",
       "      <td>Lagoas costeiras</td>\n",
       "      <td>POLYGON ((-9.22812 39.43133, -9.22765 39.43071...</td>\n",
       "      <td>521</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2609</th>\n",
       "      <td>Desembocaduras fluviais</td>\n",
       "      <td>POLYGON ((-8.99046 38.92315, -8.99021 38.92267...</td>\n",
       "      <td>522</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2610</th>\n",
       "      <td>Oceano</td>\n",
       "      <td>POLYGON ((-8.87062 41.81409, -8.87056 41.81326...</td>\n",
       "      <td>523</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>2611 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                      Legenda  \\\n",
       "0      Tecido urbano contínuo   \n",
       "1      Tecido urbano contínuo   \n",
       "2      Tecido urbano contínuo   \n",
       "3      Tecido urbano contínuo   \n",
       "4      Tecido urbano contínuo   \n",
       "...                       ...   \n",
       "2606           Planos de água   \n",
       "2607           Planos de água   \n",
       "2608         Lagoas costeiras   \n",
       "2609  Desembocaduras fluviais   \n",
       "2610                   Oceano   \n",
       "\n",
       "                                               geometry  CLC2012  \n",
       "0     POLYGON ((-9.27768 38.76588, -9.27746 38.76555...      111  \n",
       "1     POLYGON ((-9.25850 38.76457, -9.25850 38.76158...      111  \n",
       "2     POLYGON ((-9.21069 38.76698, -9.21064 38.76633...      111  \n",
       "3     POLYGON ((-9.22487 38.76998, -9.22474 38.76907...      111  \n",
       "4     POLYGON ((-9.15237 38.77167, -9.15397 38.77135...      111  \n",
       "...                                                 ...      ...  \n",
       "2606  POLYGON ((-9.30538 39.32667, -9.30438 39.32591...      512  \n",
       "2607  POLYGON ((-9.12876 39.36121, -9.12819 39.36034...      512  \n",
       "2608  POLYGON ((-9.22812 39.43133, -9.22765 39.43071...      521  \n",
       "2609  POLYGON ((-8.99046 38.92315, -8.99021 38.92267...      522  \n",
       "2610  POLYGON ((-8.87062 41.81409, -8.87056 41.81326...      523  \n",
       "\n",
       "[2611 rows x 3 columns]"
      ]
     },
     "execution_count": 194,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "shapefile"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 195,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\ProgramData\\Anaconda3\\envs\\geo_py37\\lib\\site-packages\\pyproj\\crs.py:77: FutureWarning: '+init=<authority>:<code>' syntax is deprecated. '<authority>:<code>' is the preferred initialization method.\n",
      "  return _prepare_from_string(\" \".join(pjargs))\n"
     ]
    }
   ],
   "source": [
    "shapefile = shapefile.to_crs({'init': 'epsg:32629'})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 196,
   "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>minx</th>\n",
       "      <th>miny</th>\n",
       "      <th>maxx</th>\n",
       "      <th>maxy</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>474782.521060</td>\n",
       "      <td>4.288877e+06</td>\n",
       "      <td>476430.681811</td>\n",
       "      <td>4.290843e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>476591.348965</td>\n",
       "      <td>4.289766e+06</td>\n",
       "      <td>477555.487955</td>\n",
       "      <td>4.290956e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>480560.441620</td>\n",
       "      <td>4.290757e+06</td>\n",
       "      <td>481700.696947</td>\n",
       "      <td>4.291425e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>477478.133322</td>\n",
       "      <td>4.287956e+06</td>\n",
       "      <td>482136.028108</td>\n",
       "      <td>4.291520e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>486477.998487</td>\n",
       "      <td>4.291415e+06</td>\n",
       "      <td>487108.983959</td>\n",
       "      <td>4.292233e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2606</th>\n",
       "      <td>472616.358933</td>\n",
       "      <td>4.352466e+06</td>\n",
       "      <td>474147.798607</td>\n",
       "      <td>4.353889e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2607</th>\n",
       "      <td>488111.339739</td>\n",
       "      <td>4.355853e+06</td>\n",
       "      <td>490649.235019</td>\n",
       "      <td>4.357006e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2608</th>\n",
       "      <td>479637.064347</td>\n",
       "      <td>4.359617e+06</td>\n",
       "      <td>484220.218501</td>\n",
       "      <td>4.364724e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2609</th>\n",
       "      <td>471843.279019</td>\n",
       "      <td>4.272805e+06</td>\n",
       "      <td>507900.408025</td>\n",
       "      <td>4.309807e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2610</th>\n",
       "      <td>431145.184713</td>\n",
       "      <td>4.065596e+06</td>\n",
       "      <td>659593.163080</td>\n",
       "      <td>4.636431e+06</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>2611 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "               minx          miny           maxx          maxy\n",
       "0     474782.521060  4.288877e+06  476430.681811  4.290843e+06\n",
       "1     476591.348965  4.289766e+06  477555.487955  4.290956e+06\n",
       "2     480560.441620  4.290757e+06  481700.696947  4.291425e+06\n",
       "3     477478.133322  4.287956e+06  482136.028108  4.291520e+06\n",
       "4     486477.998487  4.291415e+06  487108.983959  4.292233e+06\n",
       "...             ...           ...            ...           ...\n",
       "2606  472616.358933  4.352466e+06  474147.798607  4.353889e+06\n",
       "2607  488111.339739  4.355853e+06  490649.235019  4.357006e+06\n",
       "2608  479637.064347  4.359617e+06  484220.218501  4.364724e+06\n",
       "2609  471843.279019  4.272805e+06  507900.408025  4.309807e+06\n",
       "2610  431145.184713  4.065596e+06  659593.163080  4.636431e+06\n",
       "\n",
       "[2611 rows x 4 columns]"
      ]
     },
     "execution_count": 196,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "shapefile.bounds"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Img_data_1 cutted"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 197,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Proj('+proj=utm +zone=29 +datum=WGS84 +units=m +no_defs', preserve_units=True)\n",
      "raster bounds in current crs :\n",
      " ((464960.0, 4334040.0), (469960.0, 4334040.0), (469960.0, 4329040.0), (464960.0, 4329040.0))\n",
      "Proj('+proj=utm +zone=29 +datum=WGS84 +units=m +no_defs', preserve_units=True)\n",
      "raster bounds in current crs :\n",
      " ((459960.0, 4332040.0), (464960.0, 4332040.0), (464960.0, 4327040.0), (459960.0, 4327040.0))\n"
     ]
    }
   ],
   "source": [
    "from pyproj import transform\n",
    "from pyproj import Proj\n",
    "def cut_bounds( xmin , xmax ,  ymin , ymax ):\n",
    "    # this will get our four corner points\n",
    "    xmin = xmin \n",
    "    ymin = ymin\n",
    "    xmax = xmax\n",
    "    ymax = ymax\n",
    "    raster_gps_points = full_dataset.transform * (xmin, ymin),full_dataset.transform * (xmax, ymin),full_dataset.transform * (xmax, ymax), full_dataset.transform * (xmin, ymax),\n",
    "    # Project all longitudes, latitudes using the pyproj package\n",
    "    p1 = Proj(dataset.crs) # our current crs\n",
    "    print(p1)\n",
    "    print('raster bounds in current crs :\\n', raster_gps_points)\n",
    "    return raster_gps_points\n",
    "raster_gps_points = cut_bounds( xmin , xmax ,  ymin , ymax )\n",
    "raster_gps_points_test = cut_bounds( xmin_test , xmax_test ,  ymin_test , ymax_test )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 198,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Polygon raster gps points POLYGON ((464960 4334040, 469960 4334040, 469960 4329040, 464960 4329040, 464960 4334040))\n",
      "Imaged cutted bounds        minx       miny      maxx       maxy\n",
      "0 -9.405542  39.109805 -9.347453  39.155046\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\ProgramData\\Anaconda3\\envs\\geo_py37\\lib\\site-packages\\pyproj\\crs.py:77: FutureWarning: '+init=<authority>:<code>' syntax is deprecated. '<authority>:<code>' is the preferred initialization method.\n",
      "  return _prepare_from_string(\" \".join(pjargs))\n",
      "C:\\ProgramData\\Anaconda3\\envs\\geo_py37\\lib\\site-packages\\pyproj\\crs.py:77: FutureWarning: '+init=<authority>:<code>' syntax is deprecated. '<authority>:<code>' is the preferred initialization method.\n",
      "  return _prepare_from_string(\" \".join(pjargs))\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Shapefile bounds           minx       miny      maxx       maxy\n",
      "0    -9.290231  38.748262 -9.271269  38.765971\n",
      "1    -9.269420  38.756289 -9.258351  38.767022\n",
      "2    -9.223762  38.765308 -9.210637  38.771348\n",
      "3    -9.259214  38.740093 -9.205557  38.772182\n",
      "4    -9.155668  38.771354 -9.148402  38.778720\n",
      "...        ...        ...       ...        ...\n",
      "2606 -9.317734  39.321207 -9.299920  39.333992\n",
      "2607 -9.137997  39.352063 -9.108531  39.362436\n",
      "2608 -9.236594  39.385820 -9.183275  39.431828\n",
      "2609 -9.323687  38.603732 -8.908901  38.937196\n",
      "2610 -9.792798  36.730845 -7.206483  41.879798\n",
      "\n",
      "[2611 rows x 4 columns]\n",
      "Intersected df     index                                           Legenda  \\\n",
      "0      33                            Tecido urbano contínuo   \n",
      "1     241                         Tecido urbano descontínuo   \n",
      "2     247                         Tecido urbano descontínuo   \n",
      "3     255                         Tecido urbano descontínuo   \n",
      "4     669                  Culturas temporárias de sequeiro   \n",
      "5     679                  Culturas temporárias de sequeiro   \n",
      "6     685                  Culturas temporárias de sequeiro   \n",
      "7     686                  Culturas temporárias de sequeiro   \n",
      "8     688                  Culturas temporárias de sequeiro   \n",
      "9     742                   Culturas temporárias de regadio   \n",
      "10   1122         Sistemas culturais e parcelares complexos   \n",
      "11   1123         Sistemas culturais e parcelares complexos   \n",
      "12   1125         Sistemas culturais e parcelares complexos   \n",
      "13   1131         Sistemas culturais e parcelares complexos   \n",
      "14   1134         Sistemas culturais e parcelares complexos   \n",
      "15   1136         Sistemas culturais e parcelares complexos   \n",
      "16   1213         Sistemas culturais e parcelares complexos   \n",
      "17   1494  Agricultura com espaços naturais e semi-naturais   \n",
      "18   1496  Agricultura com espaços naturais e semi-naturais   \n",
      "19   1501  Agricultura com espaços naturais e semi-naturais   \n",
      "20   1508  Agricultura com espaços naturais e semi-naturais   \n",
      "21   1517  Agricultura com espaços naturais e semi-naturais   \n",
      "22   1534  Agricultura com espaços naturais e semi-naturais   \n",
      "23   1849                             Florestas de folhosas   \n",
      "24   1851                             Florestas de folhosas   \n",
      "25   1853                             Florestas de folhosas   \n",
      "26   1982                            Florestas de resinosas   \n",
      "27   2329                                             Matos   \n",
      "28   2574                            Praias, dunas e areais   \n",
      "29   2610                                            Oceano   \n",
      "\n",
      "                                             geometry  CLC2012  index_right  \n",
      "0   POLYGON ((-9.38224 39.13610, -9.38118 39.13594...      111            0  \n",
      "1   POLYGON ((-9.38625 39.11781, -9.38522 39.11760...      112            0  \n",
      "2   POLYGON ((-9.35822 39.13399, -9.35783 39.13381...      112            0  \n",
      "3   POLYGON ((-9.35573 39.14784, -9.35551 39.14782...      112            0  \n",
      "4   POLYGON ((-9.37935 39.11305, -9.37890 39.11288...      211            0  \n",
      "5   POLYGON ((-9.35124 39.13688, -9.35121 39.13686...      211            0  \n",
      "6   POLYGON ((-9.36537 39.15291, -9.36496 39.15288...      211            0  \n",
      "7   POLYGON ((-9.34311 39.15612, -9.34257 39.15613...      211            0  \n",
      "8   POLYGON ((-9.35380 39.16209, -9.35330 39.16198...      211            0  \n",
      "9   POLYGON ((-9.37202 39.10982, -9.37161 39.10917...      212            0  \n",
      "10  POLYGON ((-9.38501 39.11545, -9.38499 39.11440...      242            0  \n",
      "11  POLYGON ((-9.35314 39.11344, -9.35279 39.11344...      242            0  \n",
      "12  POLYGON ((-9.36272 39.11670, -9.36118 39.11628...      242            0  \n",
      "13  POLYGON ((-9.38413 39.13054, -9.38392 39.13049...      242            0  \n",
      "14  POLYGON ((-9.36446 39.13380, -9.36495 39.13267...      242            0  \n",
      "15  POLYGON ((-9.35627 39.13557, -9.35540 39.13552...      242            0  \n",
      "16  POLYGON ((-9.30738 39.35661, -9.30670 39.35561...      242            0  \n",
      "17  POLYGON ((-9.36601 39.10919, -9.36548 39.10906...      243            0  \n",
      "18  POLYGON ((-9.34274 39.11982, -9.34274 39.11957...      243            0  \n",
      "19  POLYGON ((-9.35910 39.12056, -9.35974 39.12054...      243            0  \n",
      "20  POLYGON ((-9.36473 39.12308, -9.36510 39.12282...      243            0  \n",
      "21  POLYGON ((-9.35489 39.14096, -9.35438 39.14057...      243            0  \n",
      "22  POLYGON ((-9.35534 39.16173, -9.35678 39.16154...      243            0  \n",
      "23  POLYGON ((-9.35191 39.13558, -9.35121 39.13542...      311            0  \n",
      "24  POLYGON ((-9.36330 39.13775, -9.36253 39.13774...      311            0  \n",
      "25  POLYGON ((-9.34490 39.14105, -9.34512 39.14074...      311            0  \n",
      "26  POLYGON ((-9.36825 39.13473, -9.36794 39.13465...      312            0  \n",
      "27  POLYGON ((-9.35972 39.16864, -9.35919 39.16839...      322            0  \n",
      "28  POLYGON ((-9.38918 39.10627, -9.38878 39.10558...      331            0  \n",
      "29  POLYGON ((-8.87062 41.81409, -8.87056 41.81326...      523            0  \n",
      "Intersected crs {'init': 'epsg:4326', 'no_defs': True}\n",
      "Intersected bounds         minx       miny      maxx       maxy\n",
      "0  -9.385348  39.129342 -9.373846  39.136129\n",
      "1  -9.393315  39.112158 -9.382675  39.118861\n",
      "2  -9.382986  39.107055 -9.355782  39.134736\n",
      "3  -9.382242  39.132372 -9.350183  39.151957\n",
      "4  -9.386695  39.104219 -9.372128  39.116152\n",
      "5  -9.361471  39.113438 -9.339047  39.137040\n",
      "6  -9.371393  39.147011 -9.362897  39.154949\n",
      "7  -9.347818  39.146418 -9.328153  39.156131\n",
      "8  -9.360660  39.151547 -9.351045  39.168980\n",
      "9  -9.377607  39.102357 -9.369505  39.109984\n",
      "10 -9.424319  39.009384 -9.352149  39.115720\n",
      "11 -9.364026  39.089601 -9.344767  39.117164\n",
      "12 -9.370022  39.113350 -9.358913  39.120745\n",
      "13 -9.393211  39.112879 -9.374702  39.130599\n",
      "14 -9.378663  39.120338 -9.359697  39.134496\n",
      "15 -9.359853  39.120707 -9.349516  39.135575\n",
      "16 -9.364021  39.123964 -9.158486  39.361480\n",
      "17 -9.369590  39.102890 -9.359325  39.111293\n",
      "18 -9.351651  39.109880 -9.342735  39.120009\n",
      "19 -9.365084  39.116280 -9.357687  39.123829\n",
      "20 -9.370496  39.121644 -9.362913  39.129738\n",
      "21 -9.365992  39.131448 -9.354380  39.146384\n",
      "22 -9.364819  39.146342 -9.355345  39.166611\n",
      "23 -9.355400  39.130384 -9.348698  39.136048\n",
      "24 -9.368451  39.131285 -9.360981  39.139984\n",
      "25 -9.356132  39.130794 -9.338764  39.142884\n",
      "26 -9.375123  39.132897 -9.365951  39.141634\n",
      "27 -9.368191  39.152884 -9.357831  39.168932\n",
      "28 -9.402329  39.100989 -9.388246  39.113455\n",
      "29 -9.792798  36.730845 -7.206483  41.879798\n"
     ]
    }
   ],
   "source": [
    "def shapefile_from_rasterbounds(raster_gps_points,shapefile) :\n",
    "    polygon = Polygon( list(raster_gps_points))\n",
    "    print(\"Polygon raster gps points\",polygon)\n",
    "    imageBounds = gp.GeoDataFrame(crs=CRS.from_epsg(32629))\n",
    "    imageBounds.geometry = [polygon]\n",
    "    imageBounds = imageBounds.to_crs(fiona.crs.from_epsg(4326))\n",
    "    print(\"Imaged cutted bounds\",imageBounds.bounds)\n",
    "    shapefile = shapefile.to_crs(fiona.crs.from_epsg(4326))\n",
    "    print(\"Shapefile bounds\",shapefile.bounds)\n",
    "    intersected = gp.sjoin( shapefile, imageBounds, how='inner', op='intersects', lsuffix='left', rsuffix='right')\n",
    "    intersected = intersected.reset_index()\n",
    "    print(\"Intersected df\",intersected)\n",
    "    print(\"Intersected crs\",intersected.crs)\n",
    "    print(\"Intersected bounds\",intersected.bounds)\n",
    "    return intersected\n",
    "intersected = shapefile_from_rasterbounds(raster_gps_points,shapefile)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 199,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Polygon raster gps points POLYGON ((459960 4332040, 464960 4332040, 464960 4327040, 459960 4327040, 459960 4332040))\n",
      "Imaged cutted bounds        minx       miny      maxx       maxy\n",
      "0 -9.463291  39.091567 -9.405181  39.136837\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\ProgramData\\Anaconda3\\envs\\geo_py37\\lib\\site-packages\\pyproj\\crs.py:77: FutureWarning: '+init=<authority>:<code>' syntax is deprecated. '<authority>:<code>' is the preferred initialization method.\n",
      "  return _prepare_from_string(\" \".join(pjargs))\n",
      "C:\\ProgramData\\Anaconda3\\envs\\geo_py37\\lib\\site-packages\\pyproj\\crs.py:77: FutureWarning: '+init=<authority>:<code>' syntax is deprecated. '<authority>:<code>' is the preferred initialization method.\n",
      "  return _prepare_from_string(\" \".join(pjargs))\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Shapefile bounds           minx       miny      maxx       maxy\n",
      "0    -9.290231  38.748262 -9.271269  38.765971\n",
      "1    -9.269420  38.756289 -9.258351  38.767022\n",
      "2    -9.223762  38.765308 -9.210637  38.771348\n",
      "3    -9.259214  38.740093 -9.205557  38.772182\n",
      "4    -9.155668  38.771354 -9.148402  38.778720\n",
      "...        ...        ...       ...        ...\n",
      "2606 -9.317734  39.321207 -9.299920  39.333992\n",
      "2607 -9.137997  39.352063 -9.108531  39.362436\n",
      "2608 -9.236594  39.385820 -9.183275  39.431828\n",
      "2609 -9.323687  38.603732 -8.908901  38.937196\n",
      "2610 -9.792798  36.730845 -7.206483  41.879798\n",
      "\n",
      "[2611 rows x 4 columns]\n",
      "Intersected df    index                                    Legenda  \\\n",
      "0   1122  Sistemas culturais e parcelares complexos   \n",
      "1   2610                                     Oceano   \n",
      "\n",
      "                                            geometry  CLC2012  index_right  \n",
      "0  POLYGON ((-9.38501 39.11545, -9.38499 39.11440...      242            0  \n",
      "1  POLYGON ((-8.87062 41.81409, -8.87056 41.81326...      523            0  \n",
      "Intersected crs {'init': 'epsg:4326', 'no_defs': True}\n",
      "Intersected bounds        minx       miny      maxx       maxy\n",
      "0 -9.424319  39.009384 -9.352149  39.115720\n",
      "1 -9.792798  36.730845 -7.206483  41.879798\n"
     ]
    }
   ],
   "source": [
    "intersected_test = shapefile_from_rasterbounds(raster_gps_points_test,shapefile)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 200,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "111\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "'Territórios artificializados'"
      ]
     },
     "execution_count": 200,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import re\n",
    "def class_from_CLCcode(clc) :\n",
    "    mega_classes = {\n",
    "        '^1.*' : 'Territórios artificializados',\n",
    "        '^2.*' : 'Agricultura',\n",
    "        '^31.*' : 'Floresta',\n",
    "        '^3.4' : 'Floresta',\n",
    "        '^32[123]' : 'Vegetação natural',\n",
    "        '^33[123]' : 'Espaços descobertos ou com vegetação esparsa',\n",
    "        '^41.*' : 'Zonas húmidas',\n",
    "        '^42[13]' : 'Zonas húmidas',\n",
    "        '^5.*' : 'Corpos de água',\n",
    "        '^422' : 'Corpos de água'\n",
    "    }\n",
    "    for mega_class_key in mega_classes :\n",
    "        searched =  re.search(mega_class_key, str(clc))\n",
    "        if searched is not None:\n",
    "            return mega_classes[mega_class_key]\n",
    "    return 'Not Defined'\n",
    "\n",
    "x = intersected['CLC2012'][0] \n",
    "print (x)\n",
    "class_from_CLCcode (x) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 201,
   "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>index</th>\n",
       "      <th>Legenda</th>\n",
       "      <th>geometry</th>\n",
       "      <th>CLC2012</th>\n",
       "      <th>index_right</th>\n",
       "      <th>Mega_Legenda</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>33</td>\n",
       "      <td>Tecido urbano contínuo</td>\n",
       "      <td>POLYGON ((-9.38224 39.13610, -9.38118 39.13594...</td>\n",
       "      <td>111</td>\n",
       "      <td>0</td>\n",
       "      <td>Territórios artificializados</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>241</td>\n",
       "      <td>Tecido urbano descontínuo</td>\n",
       "      <td>POLYGON ((-9.38625 39.11781, -9.38522 39.11760...</td>\n",
       "      <td>112</td>\n",
       "      <td>0</td>\n",
       "      <td>Territórios artificializados</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>247</td>\n",
       "      <td>Tecido urbano descontínuo</td>\n",
       "      <td>POLYGON ((-9.35822 39.13399, -9.35783 39.13381...</td>\n",
       "      <td>112</td>\n",
       "      <td>0</td>\n",
       "      <td>Territórios artificializados</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>255</td>\n",
       "      <td>Tecido urbano descontínuo</td>\n",
       "      <td>POLYGON ((-9.35573 39.14784, -9.35551 39.14782...</td>\n",
       "      <td>112</td>\n",
       "      <td>0</td>\n",
       "      <td>Territórios artificializados</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>669</td>\n",
       "      <td>Culturas temporárias de sequeiro</td>\n",
       "      <td>POLYGON ((-9.37935 39.11305, -9.37890 39.11288...</td>\n",
       "      <td>211</td>\n",
       "      <td>0</td>\n",
       "      <td>Agricultura</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   index                           Legenda  \\\n",
       "0     33            Tecido urbano contínuo   \n",
       "1    241         Tecido urbano descontínuo   \n",
       "2    247         Tecido urbano descontínuo   \n",
       "3    255         Tecido urbano descontínuo   \n",
       "4    669  Culturas temporárias de sequeiro   \n",
       "\n",
       "                                            geometry  CLC2012  index_right  \\\n",
       "0  POLYGON ((-9.38224 39.13610, -9.38118 39.13594...      111            0   \n",
       "1  POLYGON ((-9.38625 39.11781, -9.38522 39.11760...      112            0   \n",
       "2  POLYGON ((-9.35822 39.13399, -9.35783 39.13381...      112            0   \n",
       "3  POLYGON ((-9.35573 39.14784, -9.35551 39.14782...      112            0   \n",
       "4  POLYGON ((-9.37935 39.11305, -9.37890 39.11288...      211            0   \n",
       "\n",
       "                   Mega_Legenda  \n",
       "0  Territórios artificializados  \n",
       "1  Territórios artificializados  \n",
       "2  Territórios artificializados  \n",
       "3  Territórios artificializados  \n",
       "4                   Agricultura  "
      ]
     },
     "execution_count": 201,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "for i in intersected.index:\n",
    "    clc = intersected.at[i, 'CLC2012']\n",
    "    new_class = class_from_CLCcode(clc)\n",
    "    intersected.at[i, 'Mega_Legenda'] = new_class\n",
    "intersected.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 202,
   "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>index</th>\n",
       "      <th>Legenda</th>\n",
       "      <th>geometry</th>\n",
       "      <th>CLC2012</th>\n",
       "      <th>index_right</th>\n",
       "      <th>Mega_Legenda</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1122</td>\n",
       "      <td>Sistemas culturais e parcelares complexos</td>\n",
       "      <td>POLYGON ((-9.38501 39.11545, -9.38499 39.11440...</td>\n",
       "      <td>242</td>\n",
       "      <td>0</td>\n",
       "      <td>Agricultura</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2610</td>\n",
       "      <td>Oceano</td>\n",
       "      <td>POLYGON ((-8.87062 41.81409, -8.87056 41.81326...</td>\n",
       "      <td>523</td>\n",
       "      <td>0</td>\n",
       "      <td>Corpos de água</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   index                                    Legenda  \\\n",
       "0   1122  Sistemas culturais e parcelares complexos   \n",
       "1   2610                                     Oceano   \n",
       "\n",
       "                                            geometry  CLC2012  index_right  \\\n",
       "0  POLYGON ((-9.38501 39.11545, -9.38499 39.11440...      242            0   \n",
       "1  POLYGON ((-8.87062 41.81409, -8.87056 41.81326...      523            0   \n",
       "\n",
       "     Mega_Legenda  \n",
       "0     Agricultura  \n",
       "1  Corpos de água  "
      ]
     },
     "execution_count": 202,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "for i in intersected_test.index:\n",
    "    clc = intersected_test.at[i, 'CLC2012']\n",
    "    new_class = class_from_CLCcode(clc)\n",
    "    intersected_test.at[i, 'Mega_Legenda'] = new_class\n",
    "intersected_test.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 203,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of different classes :6\n",
      "['Agricultura' 'Corpos de água'\n",
      " 'Espaços descobertos ou com vegetação esparsa' 'Floresta'\n",
      " 'Territórios artificializados' 'Vegetação natural']\n"
     ]
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "unique, counts = np.unique(intersected['Mega_Legenda'], return_counts=True)\n",
    "plt.bar(unique, counts, 1)\n",
    "plt.title('Class Frequency')\n",
    "plt.xlabel('Class')\n",
    "plt.ylabel('Frequency')\n",
    "plt.show()\n",
    "\n",
    "print('Number of different classes :'  + str(len(unique)))\n",
    "print(unique)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Setting up label data with bands"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 204,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Data crs: EPSG:32629\n",
      "Before crs: {'init': 'epsg:4326', 'no_defs': True}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\ProgramData\\Anaconda3\\envs\\geo_py37\\lib\\site-packages\\pyproj\\crs.py:77: FutureWarning: '+init=<authority>:<code>' syntax is deprecated. '<authority>:<code>' is the preferred initialization method.\n",
      "  return _prepare_from_string(\" \".join(pjargs))\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "After crs: {'init': 'epsg:32629'}\n",
      "Data crs: EPSG:32629\n",
      "Before crs: {'init': 'epsg:4326', 'no_defs': True}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\ProgramData\\Anaconda3\\envs\\geo_py37\\lib\\site-packages\\pyproj\\crs.py:77: FutureWarning: '+init=<authority>:<code>' syntax is deprecated. '<authority>:<code>' is the preferred initialization method.\n",
      "  return _prepare_from_string(\" \".join(pjargs))\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "After crs: {'init': 'epsg:32629'}\n"
     ]
    }
   ],
   "source": [
    "def setup_shafile_to_data(shapefile,full_dataset) :\n",
    "    shapefile = shapefile.filter(['Legenda','Mega_Legenda', 'geometry','CLC2012'])\n",
    "    print(\"Data crs: \" + str(full_dataset.crs))\n",
    "    print(\"Before crs: \" + str(shapefile.crs))\n",
    "    if shapefile.crs != {'init': 'epsg:32629'} :\n",
    "        shapefile = shapefile.to_crs({'init': 'epsg:32629'})\n",
    "    print(\"After crs: \" + str(shapefile.crs))\n",
    "    return shapefile\n",
    "shapefile = setup_shafile_to_data(intersected,full_dataset)\n",
    "shapefile_test = setup_shafile_to_data(intersected_test,full_dataset)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 205,
   "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>Legenda</th>\n",
       "      <th>Mega_Legenda</th>\n",
       "      <th>geometry</th>\n",
       "      <th>CLC2012</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Tecido urbano contínuo</td>\n",
       "      <td>Territórios artificializados</td>\n",
       "      <td>POLYGON ((466964.424 4331949.479, 467056.111 4...</td>\n",
       "      <td>111</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Tecido urbano descontínuo</td>\n",
       "      <td>Territórios artificializados</td>\n",
       "      <td>POLYGON ((466609.165 4329920.686, 466698.172 4...</td>\n",
       "      <td>112</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Tecido urbano descontínuo</td>\n",
       "      <td>Territórios artificializados</td>\n",
       "      <td>POLYGON ((469039.730 4331706.460, 469073.026 4...</td>\n",
       "      <td>112</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Tecido urbano descontínuo</td>\n",
       "      <td>Territórios artificializados</td>\n",
       "      <td>POLYGON ((469260.812 4333243.314, 469280.171 4...</td>\n",
       "      <td>112</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Culturas temporárias de sequeiro</td>\n",
       "      <td>Agricultura</td>\n",
       "      <td>POLYGON ((467203.867 4329390.271, 467242.736 4...</td>\n",
       "      <td>211</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                            Legenda                  Mega_Legenda  \\\n",
       "0            Tecido urbano contínuo  Territórios artificializados   \n",
       "1         Tecido urbano descontínuo  Territórios artificializados   \n",
       "2         Tecido urbano descontínuo  Territórios artificializados   \n",
       "3         Tecido urbano descontínuo  Territórios artificializados   \n",
       "4  Culturas temporárias de sequeiro                   Agricultura   \n",
       "\n",
       "                                            geometry  CLC2012  \n",
       "0  POLYGON ((466964.424 4331949.479, 467056.111 4...      111  \n",
       "1  POLYGON ((466609.165 4329920.686, 466698.172 4...      112  \n",
       "2  POLYGON ((469039.730 4331706.460, 469073.026 4...      112  \n",
       "3  POLYGON ((469260.812 4333243.314, 469280.171 4...      112  \n",
       "4  POLYGON ((467203.867 4329390.271, 467242.736 4...      211  "
      ]
     },
     "execution_count": 205,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "shapefile.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Choose target column = Legenda or Mega_Legenda"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 206,
   "metadata": {},
   "outputs": [],
   "source": [
    "Class = 'Mega_Legenda'\n",
    "#Class = 'Legenda' "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 207,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "List of Land Cover Classes:\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "array(['Agricultura', 'Corpos de água',\n",
       "       'Espaços descobertos ou com vegetação esparsa', 'Floresta',\n",
       "       'Territórios artificializados', 'Vegetação natural'], dtype=object)"
      ]
     },
     "execution_count": 207,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "unique= np.unique(shapefile[Class])\n",
    "print('List of Land Cover Classes:')\n",
    "unique"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 208,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "List of Land Cover Classes for test:\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "array(['Agricultura', 'Corpos de água'], dtype=object)"
      ]
     },
     "execution_count": 208,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "unique= np.unique(shapefile_test[Class])\n",
    "print('List of Land Cover Classes for test:')\n",
    "unique"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 209,
   "metadata": {},
   "outputs": [],
   "source": [
    "#shapefile.geometry = shapefile.geometry.map(lambda polygon: shapely.ops.transform(lambda x, y: (y, x), polygon))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 210,
   "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>minx</th>\n",
       "      <th>miny</th>\n",
       "      <th>maxx</th>\n",
       "      <th>maxy</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>466693.388761</td>\n",
       "      <td>4.331197e+06</td>\n",
       "      <td>467687.138734</td>\n",
       "      <td>4.331952e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>465996.598721</td>\n",
       "      <td>4.329296e+06</td>\n",
       "      <td>466917.843856</td>\n",
       "      <td>4.330039e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>466894.548139</td>\n",
       "      <td>4.328722e+06</td>\n",
       "      <td>469242.378114</td>\n",
       "      <td>4.331790e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>466964.423703</td>\n",
       "      <td>4.331532e+06</td>\n",
       "      <td>469737.277794</td>\n",
       "      <td>4.333700e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>466568.437188</td>\n",
       "      <td>4.328411e+06</td>\n",
       "      <td>467826.714693</td>\n",
       "      <td>4.329736e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>468752.564988</td>\n",
       "      <td>4.329424e+06</td>\n",
       "      <td>470697.015567</td>\n",
       "      <td>4.332043e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>467908.277434</td>\n",
       "      <td>4.333156e+06</td>\n",
       "      <td>468642.157173</td>\n",
       "      <td>4.334036e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>469946.590040</td>\n",
       "      <td>4.333078e+06</td>\n",
       "      <td>471644.251313</td>\n",
       "      <td>4.334159e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>468840.714745</td>\n",
       "      <td>4.333654e+06</td>\n",
       "      <td>469671.264515</td>\n",
       "      <td>4.335590e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>467350.703474</td>\n",
       "      <td>4.328203e+06</td>\n",
       "      <td>468052.143352</td>\n",
       "      <td>4.329048e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>463267.162078</td>\n",
       "      <td>4.317901e+06</td>\n",
       "      <td>469533.727742</td>\n",
       "      <td>4.329689e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>468524.038318</td>\n",
       "      <td>4.326780e+06</td>\n",
       "      <td>470193.392006</td>\n",
       "      <td>4.329838e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>468012.400858</td>\n",
       "      <td>4.329418e+06</td>\n",
       "      <td>468971.638860</td>\n",
       "      <td>4.330240e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>466008.818077</td>\n",
       "      <td>4.329371e+06</td>\n",
       "      <td>467606.508072</td>\n",
       "      <td>4.331340e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>467269.269603</td>\n",
       "      <td>4.330197e+06</td>\n",
       "      <td>468910.063616</td>\n",
       "      <td>4.331765e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>468894.740277</td>\n",
       "      <td>4.330232e+06</td>\n",
       "      <td>469787.383191</td>\n",
       "      <td>4.331882e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>468634.432655</td>\n",
       "      <td>4.330579e+06</td>\n",
       "      <td>486325.259249</td>\n",
       "      <td>4.356937e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>468044.827398</td>\n",
       "      <td>4.328258e+06</td>\n",
       "      <td>468931.634532</td>\n",
       "      <td>4.329192e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>469597.692009</td>\n",
       "      <td>4.329029e+06</td>\n",
       "      <td>470371.910787</td>\n",
       "      <td>4.330150e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>468439.544499</td>\n",
       "      <td>4.329743e+06</td>\n",
       "      <td>469079.903020</td>\n",
       "      <td>4.330581e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>467974.271246</td>\n",
       "      <td>4.330341e+06</td>\n",
       "      <td>468629.919600</td>\n",
       "      <td>4.331238e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>468370.401979</td>\n",
       "      <td>4.331426e+06</td>\n",
       "      <td>469374.417859</td>\n",
       "      <td>4.333082e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>468479.844679</td>\n",
       "      <td>4.333078e+06</td>\n",
       "      <td>469300.202943</td>\n",
       "      <td>4.335327e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>469283.317546</td>\n",
       "      <td>4.331305e+06</td>\n",
       "      <td>469861.870661</td>\n",
       "      <td>4.331934e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>468155.947045</td>\n",
       "      <td>4.331409e+06</td>\n",
       "      <td>468800.522222</td>\n",
       "      <td>4.332375e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>469221.384723</td>\n",
       "      <td>4.331348e+06</td>\n",
       "      <td>470722.607067</td>\n",
       "      <td>4.332690e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>467578.643139</td>\n",
       "      <td>4.331591e+06</td>\n",
       "      <td>468371.630292</td>\n",
       "      <td>4.332559e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>468187.343730</td>\n",
       "      <td>4.333806e+06</td>\n",
       "      <td>469088.339943</td>\n",
       "      <td>4.335585e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>465211.175885</td>\n",
       "      <td>4.328061e+06</td>\n",
       "      <td>466430.275679</td>\n",
       "      <td>4.329441e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>431145.184713</td>\n",
       "      <td>4.065596e+06</td>\n",
       "      <td>659593.163080</td>\n",
       "      <td>4.636431e+06</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             minx          miny           maxx          maxy\n",
       "0   466693.388761  4.331197e+06  467687.138734  4.331952e+06\n",
       "1   465996.598721  4.329296e+06  466917.843856  4.330039e+06\n",
       "2   466894.548139  4.328722e+06  469242.378114  4.331790e+06\n",
       "3   466964.423703  4.331532e+06  469737.277794  4.333700e+06\n",
       "4   466568.437188  4.328411e+06  467826.714693  4.329736e+06\n",
       "5   468752.564988  4.329424e+06  470697.015567  4.332043e+06\n",
       "6   467908.277434  4.333156e+06  468642.157173  4.334036e+06\n",
       "7   469946.590040  4.333078e+06  471644.251313  4.334159e+06\n",
       "8   468840.714745  4.333654e+06  469671.264515  4.335590e+06\n",
       "9   467350.703474  4.328203e+06  468052.143352  4.329048e+06\n",
       "10  463267.162078  4.317901e+06  469533.727742  4.329689e+06\n",
       "11  468524.038318  4.326780e+06  470193.392006  4.329838e+06\n",
       "12  468012.400858  4.329418e+06  468971.638860  4.330240e+06\n",
       "13  466008.818077  4.329371e+06  467606.508072  4.331340e+06\n",
       "14  467269.269603  4.330197e+06  468910.063616  4.331765e+06\n",
       "15  468894.740277  4.330232e+06  469787.383191  4.331882e+06\n",
       "16  468634.432655  4.330579e+06  486325.259249  4.356937e+06\n",
       "17  468044.827398  4.328258e+06  468931.634532  4.329192e+06\n",
       "18  469597.692009  4.329029e+06  470371.910787  4.330150e+06\n",
       "19  468439.544499  4.329743e+06  469079.903020  4.330581e+06\n",
       "20  467974.271246  4.330341e+06  468629.919600  4.331238e+06\n",
       "21  468370.401979  4.331426e+06  469374.417859  4.333082e+06\n",
       "22  468479.844679  4.333078e+06  469300.202943  4.335327e+06\n",
       "23  469283.317546  4.331305e+06  469861.870661  4.331934e+06\n",
       "24  468155.947045  4.331409e+06  468800.522222  4.332375e+06\n",
       "25  469221.384723  4.331348e+06  470722.607067  4.332690e+06\n",
       "26  467578.643139  4.331591e+06  468371.630292  4.332559e+06\n",
       "27  468187.343730  4.333806e+06  469088.339943  4.335585e+06\n",
       "28  465211.175885  4.328061e+06  466430.275679  4.329441e+06\n",
       "29  431145.184713  4.065596e+06  659593.163080  4.636431e+06"
      ]
     },
     "execution_count": 210,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "shapefile.bounds"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 211,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "raster bounds in current crs :\n",
      " ((464960.0, 4334040.0), (469960.0, 4334040.0), (469960.0, 4329040.0), (464960.0, 4329040.0))\n"
     ]
    }
   ],
   "source": [
    "print('raster bounds in current crs :\\n', raster_gps_points)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 212,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'shapely.geometry.polygon.Polygon'>\n",
      "POLYGON ((466964.4237027746 4331949.479068073, 467056.111145639 4331931.712223588, 467149.0952492154 4331951.234501887, 467195.676337795 4331951.679405843, 467233.3416034218 4331910.11235441, 467266.3709926478 4331866.171847544, 467332.0516111786 4331817.884539855, 467369.6945313095 4331778.64652798, 467372.3127682726 4331748.391088762, 467391.6125745543 4331678.697532657, 467454.6972043611 4331658.336574916, 467550.6555184731 4331610.338493205, 467611.8873344366 4331536.322760579, 467595.4293370568 4331424.44890151, 467638.0257973998 4331385.875851696, 467670.3637442201 4331322.715065499, 467687.1387342545 4331227.671187854, 467672.6153339227 4331197.917508592, 467650.105950683 4331196.678844143, 467572.913497621 4331230.880677446, 467456.3941267077 4331236.755756233, 467405.0214429033 4331250.24069206, 467309.4857897328 4331253.986883694, 467197.7358053764 4331248.261140807, 467150.9991210354 4331264.119713548, 467120.4323009839 4331294.108234264, 467052.6228797645 4331321.411812863, 466987.4094343916 4331320.789023704, 466967.3508684837 4331290.069661702, 466952.9065789034 4331296.32751901, 466922.060721425 4331306.281973261, 466878.377490459 4331318.674987479, 466850.117319837 4331326.092774488, 466817.6371282637 4331331.393166715, 466816.5865481617 4331327.163780027, 466798.6228633797 4331332.96173092, 466738.0230794031 4331337.041440571, 466693.3887614207 4331340.136671946, 466695.4174970426 4331347.502681723, 466729.8309906102 4331472.445875268, 466754.6842511581 4331562.679373566, 466793.6832239885 4331675.278798358, 466884.8615193169 4331780.349090354, 466955.4990407063 4331937.338570477, 466964.4237027746 4331949.479068073))\n",
      "<class 'list'>\n",
      "[{'type': 'Polygon', 'coordinates': (((466964.42370277457, 4331949.4790680725), (467056.111145639, 4331931.712223588), (467149.0952492154, 4331951.234501887), (467195.676337795, 4331951.679405843), (467233.3416034218, 4331910.11235441), (467266.37099264783, 4331866.171847544), (467332.0516111786, 4331817.884539855), (467369.69453130953, 4331778.64652798), (467372.3127682726, 4331748.391088762), (467391.61257455434, 4331678.697532657), (467454.69720436115, 4331658.336574916), (467550.65551847307, 4331610.338493205), (467611.88733443664, 4331536.322760579), (467595.42933705676, 4331424.44890151), (467638.0257973998, 4331385.875851696), (467670.36374422006, 4331322.715065499), (467687.13873425446, 4331227.671187854), (467672.6153339227, 4331197.917508592), (467650.105950683, 4331196.678844143), (467572.91349762096, 4331230.880677446), (467456.39412670775, 4331236.755756233), (467405.02144290326, 4331250.24069206), (467309.48578973283, 4331253.986883694), (467197.73580537643, 4331248.261140807), (467150.99912103545, 4331264.119713548), (467120.4323009839, 4331294.108234264), (467052.6228797645, 4331321.411812863), (466987.4094343916, 4331320.789023704), (466967.35086848366, 4331290.069661702), (466952.90657890344, 4331296.32751901), (466922.060721425, 4331306.281973261), (466878.37749045895, 4331318.674987479), (466850.11731983704, 4331326.092774488), (466817.63712826365, 4331331.393166715), (466816.5865481617, 4331327.163780027), (466798.62286337966, 4331332.96173092), (466738.0230794031, 4331337.041440571), (466693.38876142073, 4331340.136671946), (466695.4174970426, 4331347.502681723), (466729.83099061024, 4331472.445875268), (466754.68425115815, 4331562.679373566), (466793.68322398845, 4331675.278798358), (466884.86151931685, 4331780.349090354), (466955.49904070626, 4331937.338570477), (466964.42370277457, 4331949.4790680725)),)}]\n"
     ]
    }
   ],
   "source": [
    "# this generates a list of shapely geometries\n",
    "geoms = shapefile.geometry.values \n",
    "geoms_test = shapefile_test.geometry.values \n",
    "\n",
    "# let's grab a single shapely geometry to check\n",
    "geometry = geoms[0] \n",
    "print(type(geometry))\n",
    "print(geometry)\n",
    "\n",
    "# transform to GeoJSON format\n",
    "from shapely.geometry import mapping\n",
    "feature = [mapping(geometry)] # can also do this using polygon.__geo_interface__\n",
    "print(type(feature))\n",
    "print(feature)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 213,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(4, 77, 100)"
      ]
     },
     "execution_count": 213,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "out_image, out_transform = rasterio.mask.mask(full_dataset, feature,crop=True)\n",
    "out_image.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 214,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Affine(10.0, 0.0, 466690.0,\n",
       "       0.0, -10.0, 4331960.0)"
      ]
     },
     "execution_count": 214,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "out_transform"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 218,
   "metadata": {},
   "outputs": [],
   "source": [
    "def setup_features_and_target(src,geoms,shapefile) : \n",
    "    X = np.array([], dtype=np.int8).reshape(0,4) # pixels for training\n",
    "    y = np.array([], dtype=np.string_) # labels for training\n",
    "    # extract the raster values within the polygon \n",
    "    band_count = src.count\n",
    "    for index, geom in enumerate(geoms):\n",
    "        feature = [mapping(geom)]\n",
    "\n",
    "        # the mask function returns an array of the raster pixels within this feature\n",
    "        out_image, out_transform = rasterio.mask.mask(src, feature,crop=True)\n",
    "        # eliminate all the pixels with 0 values for all 8 bands - AKA not actually part of the shapefile\n",
    "        out_image_trimmed = out_image[:,~np.all(out_image == 0, axis=0)]\n",
    "        # eliminate all the pixels with 255 values for all 8 bands - AKA not actually part of the shapefile\n",
    "        out_image_trimmed = out_image_trimmed[:,~np.all(out_image_trimmed == 255, axis=0)]\n",
    "        # reshape the array to [pixel count, bands]\n",
    "        out_image_reshaped = out_image_trimmed.reshape(-1, band_count)\n",
    "        # append the labels to the y array\n",
    "        y = np.append(y,[shapefile[Class][index]] * out_image_reshaped.shape[0]) \n",
    "        # stack the pizels onto the pixel array\n",
    "        X = np.vstack((X,out_image_reshaped))\n",
    "    return X,y"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 219,
   "metadata": {},
   "outputs": [],
   "source": [
    "src = full_dataset\n",
    "X,y = setup_features_and_target(src,geoms,shapefile)\n",
    "X_test,y_test = setup_features_and_target(src,geoms_test,shapefile_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 220,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The training data include 6 classes: ['Agricultura' 'Corpos de água'\n",
      " 'Espaços descobertos ou com vegetação esparsa' 'Floresta'\n",
      " 'Territórios artificializados' 'Vegetação natural']\n",
      "\n",
      "The training data include 2 classes: ['Agricultura' 'Corpos de água']\n",
      "\n",
      "Our X matrix is sized: (35129396, 4)\n",
      "Our y array is sized: (35129396,)\n",
      "Our X_test matrix is sized: (33224143, 4)\n",
      "Our y_test array is sized: (33224143,)\n"
     ]
    }
   ],
   "source": [
    "# What are our classification labels?\n",
    "labels = np.unique(shapefile[Class])\n",
    "print('The training data include {n} classes: {classes}\\n'.format(n=labels.size, \n",
    "                                                                classes=labels))\n",
    "labels_test = np.unique(shapefile_test[Class])\n",
    "print('The testing data include {n} classes: {classes}\\n'.format(n=labels_test.size, \n",
    "                                                                classes=labels_test))\n",
    "\n",
    "# We will need a \"X\" matrix containing our features, and a \"y\" array containing our labels\n",
    "print('Our X matrix is sized: {sz}'.format(sz=X.shape))\n",
    "print('Our y array is sized: {sz}'.format(sz=y.shape))\n",
    "\n",
    "print('Our X_test matrix is sized: {sz}'.format(sz=X_test.shape))\n",
    "print('Our y_test array is sized: {sz}'.format(sz=y_test.shape))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 221,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Band Intensities Higher Ref Subset')"
      ]
     },
     "execution_count": 221,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x576 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(1,3, figsize=[20,8])\n",
    "\n",
    "# numbers 1-4\n",
    "band_count = np.arange(1,5)\n",
    "\n",
    "classes = np.unique(y)\n",
    "for class_type in classes:\n",
    "    band_intensity = np.mean(X[y==class_type, :], axis=0)\n",
    "    ax[0].plot(band_count, band_intensity, label=class_type)\n",
    "    ax[1].plot(band_count, band_intensity, label=class_type)\n",
    "    ax[2].plot(band_count, band_intensity, label=class_type)\n",
    "# plot them as lines\n",
    "\n",
    "# Add some axis labels\n",
    "ax[0].set_xlabel('Band #')\n",
    "ax[0].set_ylabel('Reflectance Value')\n",
    "ax[1].set_ylabel('Reflectance Value')\n",
    "ax[1].set_xlabel('Band #')\n",
    "ax[2].set_ylabel('Reflectance Value')\n",
    "ax[2].set_xlabel('Band #')\n",
    "#ax[0].set_ylim(32,38)\n",
    "ax[1].set_ylim(32,38)\n",
    "ax[2].set_ylim(70,140)\n",
    "#ax.set\n",
    "ax[1].legend(loc=\"upper right\")\n",
    "# Add a title\n",
    "ax[0].set_title('Band Intensities Full Overview')\n",
    "ax[1].set_title('Band Intensities Lower Ref Subset')\n",
    "ax[2].set_title('Band Intensities Higher Ref Subset')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 222,
   "metadata": {},
   "outputs": [],
   "source": [
    "def str_class_to_int(class_array):\n",
    "    class_array[class_array == 'Territórios artificializados'] = 0\n",
    "    class_array[class_array == 'Agricultura'] = 1\n",
    "    class_array[class_array == 'Floresta'] = 2\n",
    "    class_array[class_array == 'Vegetação natural'] = 3\n",
    "    class_array[class_array == 'Espaços descobertos ou com vegetação esparsa'] = 4\n",
    "    class_array[class_array == 'Zonas húmidas'] = 5\n",
    "    class_array[class_array == 'Corpos de água'] = 6\n",
    "    return(class_array.astype(int))\n",
    "def str_class_to_int_v2(class_array,class_test):\n",
    "    uniques = np.unique(class_array)\n",
    "    i=0\n",
    "    for class_name in uniques :\n",
    "        class_array[class_array == class_name ] = i\n",
    "        class_test[class_test == class_name ] = i\n",
    "        i+=1\n",
    "    return(class_array.astype(int) , class_test.astype(int) ) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 223,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['Agricultura' 'Corpos de água'\n",
      " 'Espaços descobertos ou com vegetação esparsa' 'Floresta'\n",
      " 'Territórios artificializados' 'Vegetação natural']\n",
      "['Agricultura' 'Corpos de água']\n",
      "[0 1 2 3 4 6]\n",
      "[1 6]\n"
     ]
    }
   ],
   "source": [
    "print(np.unique(y))\n",
    "print(np.unique(y_test))\n",
    "y = str_class_to_int(y)\n",
    "y_test = str_class_to_int(y_test)\n",
    "print(np.unique(y))\n",
    "print(np.unique(y_test))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Supervised Classification Algorithm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.svm import LinearSVC\n",
    "from sklearn.gaussian_process import GaussianProcessClassifier\n",
    "from sklearn.naive_bayes import GaussianNB\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "#model = LinearSVC(C=1,verbose=1)\n",
    "model = KNeighborsClassifier()\n",
    "model.fit(X, y)\n",
    "score = model.score(X_test, y_test)\n",
    "score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "with rasterio.open(img_fp) as src:\n",
    "    # may need to reduce this image size if your kernel crashes, takes a lot of memory\n",
    "    img = src.read()[:, xmin_test : xmax_test ,  ymin_test : ymax_test ]\n",
    "\n",
    "# Take our full image and reshape into long 2d array (nrow * ncol, nband) for classification\n",
    "print(img.shape)\n",
    "reshaped_img = reshape_as_image(img)\n",
    "print(reshaped_img.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "class_prediction = model.predict(reshaped_img.reshape(-1, 4))\n",
    "before = class_prediction.shape\n",
    "# Reshape our classification map back into a 2D matrix so we can visualize it\n",
    "class_prediction = class_prediction.reshape(reshaped_img[:, :, 0].shape)\n",
    "class_prediction.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "\"\"\"\n",
    "reshaped_img.reshape(-1, 4).shape\n",
    "real_img = y_test[:250000] #fixme\n",
    "print(before)\n",
    "print(real_img.shape)\n",
    "print(y_test.shape)\n",
    "# Reshape our classification map back into a 2D matrix so we can visualize it\n",
    "real_class = real_img.reshape(reshaped_img[:, :, 0].shape) #fixme\n",
    "real_class.shape\n",
    "\"\"\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "real_class = real_class.astype(int)\n",
    "print(real_class)\n",
    "class_prediction = class_prediction.astype(int)\n",
    "class_prediction"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Visualizing results "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "    \n",
    "# find the highest pixel value in the prediction image\n",
    "n = int(np.max(class_prediction))\n",
    "\n",
    "# next setup a colormap for our map\n",
    "colors = dict((\n",
    "    (0, (60, 60, 60, 255)),     # Grey - Territórios artificializados\n",
    "    (1, (20,230,20, 255)),      # Light Green - Agricultura\n",
    "    (2, (5, 80, 5, 255)),       # Dark Green - Floresta\n",
    "    (3, (170, 200, 20, 255)),   # Yellow - Vegetação natural\n",
    "    (4, (50, 20, 0, 255)),      # Brown - Espaços descobertos ou com vegetação esparsa\n",
    "    (5, (0, 250, 250, 255)),    # Light Blue - Zonas húmidas\n",
    "    (6, (0, 0, 200, 255)),      # Dark Blue - Corpos de água\n",
    "))\n",
    "\n",
    "# Put 0 - 255 as float 0 - 1\n",
    "for k in colors:\n",
    "    v = colors[k]\n",
    "    _v = [_v / 255.0 for _v in v]\n",
    "    colors[k] = _v\n",
    "    \n",
    "index_colors = [colors[key] if key in colors else \n",
    "                (255, 255, 255, 0) for key in range(0, n+1)]\n",
    "\n",
    "cmap = plt.matplotlib.colors.ListedColormap(index_colors, 'Classification', n+1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def discrete_matshow(data):\n",
    "    #get discrete colormap\n",
    "    cmap = plt.get_cmap('RdBu', np.max(data)-np.min(data)+1)\n",
    "    # set limits .5 outside true range\n",
    "    mat = plt.matshow(data,cmap=cmap,vmin = np.min(data)-.5, vmax = np.max(data)+.5)\n",
    "    #tell the colorbar to tick at integers\n",
    "    cax = plt.colorbar(mat, ticks=np.arange(np.min(data),np.max(data)+1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "fig, axs = plt.subplots(1,2,figsize=(20,20))\n",
    "\n",
    "img_stretched = color_stretch(reshaped_img, [2, 1, 0])\n",
    "axs[0].imshow(img_stretched)\n",
    "\n",
    "prediction_plot= axs[1].imshow(class_prediction, cmap=cmap, interpolation='none')\n",
    "fig.colorbar(prediction_plot, ax=axs[1],ticks=range(0,6),fraction=0.046, pad=0.04)\n",
    "\n",
    "#real_plot = axs[2].imshow(real_class, cmap=cmap, interpolation='none')\n",
    "#fig.colorbar(real_plot, ax=axs[2],ticks=range(0,6),fraction=0.046, pad=0.04)\n",
    "\n",
    "fig.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "fig, axs = plt.subplots(1,2,figsize=(15,15))\n",
    "\n",
    "img_stretched = color_stretch(reshaped_img, [2, 1, 0])\n",
    "\n",
    "axs[0].imshow(img_stretched[0:100, 0:100])\n",
    "\n",
    "prediction_plot = axs[1].imshow(class_prediction[0:100, 0:100], cmap=cmap, interpolation='none')\n",
    "fig.colorbar(prediction_plot, ax=axs[1],ticks=range(0,4),fraction=0.046, pad=0.04)\n",
    "\n",
    "fig.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "fig, axs = plt.subplots(2,2,figsize=(15,7))\n",
    "\n",
    "reshaped_img =reshape_as_image( img )\n",
    "img_stretched = color_stretch(reshaped_img, [2, 1, 0])\n",
    "axs[0,0].set_title(\"True Color\")\n",
    "axs[0,0].imshow(img_stretched)\n",
    "\n",
    "axs[0,1].set_title(\"Class Prediction\")\n",
    "predictions_plot = axs[0,1].imshow(class_prediction, cmap=cmap, interpolation='none')\n",
    "fig.colorbar(predictions_plot, ax=axs[0,1],ticks=range(0,6))\n",
    "\n",
    "nwdi_plot = axs[1,0].imshow(ndwi, cmap=\"RdYlBu\")\n",
    "axs[1,0].set_title(\"NDWI\")\n",
    "fig.colorbar(nwdi_plot, ax=axs[1,0])\n",
    "\n",
    "ndvi_plot = axs[1,1].imshow(ndvi, cmap=\"RdYlGn\")\n",
    "axs[1,1].set_title(\"NDVI\")\n",
    "fig.colorbar(ndvi_plot, ax=axs[1,1])\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "alert() # audio alert"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "geo_py37_kernel",
   "language": "python",
   "name": "geo_py37_kernel"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}