np-array-nodata.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "d54d7f6a-2b7e-4932-8703-362d8a30a80c",
   "metadata": {},
   "outputs": [],
   "source": [
    "import rasterio\n",
    "from rasterio.enums import Compression\n",
    "import numpy as np\n",
    "\n",
    "from utils.raster.raster_metadata_helpers import DEFAULT_GTIFF_PROFILE"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "id": "4ade3998-0f2e-47d0-ab34-c34c567f2539",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'/Users/mitchellgritts/Downloads/departure-rasters/D_NRV.float32.deflate.tif'"
      ]
     },
     "execution_count": 90,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "filename = \"D_NRV\"\n",
    "dtype=\"float32\"\n",
    "nodata=0\n",
    "src_raster=f\"/Users/mitchellgritts/Downloads/departure-rasters/{filename}.tiff\"\n",
    "dst_raster=f\"/Users/mitchellgritts/Downloads/departure-rasters/{filename}.{dtype}.deflate.tif\"\n",
    "dst_raster"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "id": "dd1e2838-57d0-4f4d-a8f9-db32be38b2a5",
   "metadata": {},
   "outputs": [],
   "source": [
    "def change_raster_dtype(src_raster, dst_raster, dtype, nodata):\n",
    "    dtype = np.dtype(dtype)\n",
    "    with rasterio.open(src_raster, \"r\") as src:\n",
    "        src_profile = src.profile\n",
    "        src_data = src.read(1, masked=True)\n",
    "        \n",
    "    data = src_data.copy()\n",
    "    data.set_fill_value(nodata)\n",
    "    data = data.filled()\n",
    "    data = data.astype(dtype)\n",
    "    \n",
    "    profile = {**src_profile, **DEFAULT_GTIFF_PROFILE, **{\"nodata\": nodata, \"dtype\": dtype, \"compress\": Compression.deflate}}\n",
    "    print(profile)\n",
    "    \n",
    "    with rasterio.open(dst_raster, \"w\", **profile) as dst:\n",
    "        dst.write(data, 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "id": "0fc5ad4c-af59-4495-805d-e3e146538d09",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'driver': 'GTiff', 'dtype': dtype('float32'), 'nodata': 0, 'width': 22257, 'height': 22668, 'count': 1, 'crs': CRS.from_epsg(3310), 'transform': Affine(5.0, 0.0, -100625.739272266,\n",
      "       0.0, -5.0, 190467.673104919), 'blockysize': 512, 'tiled': True, 'interleave': 'pixel', 'blockxsize': 512, 'compress': <Compression.deflate: 'DEFLATE'>}\n"
     ]
    }
   ],
   "source": [
    "change_raster_dtype(src_raster, dst_raster, dtype, nodata)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3c349727-b3a8-4900-a997-5cbe6a7ab282",
   "metadata": {},
   "source": [
    "## test compression equality"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "id": "f7a590dd-51a4-4dfc-98d9-9b495c91acb2",
   "metadata": {},
   "outputs": [],
   "source": [
    "with rasterio.open(src_raster, \"r\") as src:\n",
    "    original_data = src.read(1, masked=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "id": "e8e8e0a0-733c-4d8e-a71c-a17925d9a17a",
   "metadata": {},
   "outputs": [],
   "source": [
    "with rasterio.open(dst_raster, \"r\") as src:\n",
    "    compressed_data = src.read(1, masked=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "id": "5a2dd370-3d4b-4030-b21a-a9a3349c0c9c",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(22668, 22257)\n",
      "(22668, 22257)\n"
     ]
    }
   ],
   "source": [
    "print(original_data.shape)\n",
    "print(compressed_data.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "id": "4875a834-4498-4f22-8d47-d94998ce3ccb",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.array_equal(original_data.data, compressed_data.data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "26918d05-5b22-4a60-a53e-e1e0bd5226b1",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.10.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}