{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Init"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Import Triskele\n",
    "\n",
    "I did not install triskele-python on my system folders, so this is a trick to load the module outside the triskele project."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import sys\n",
    "sys.path.append('../triskele/python/') # Change with your triskele install path\n",
    "import triskele"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Simple test\n",
    "\n",
    "Just to test the python import and the triskele compilation. Create an array with random values."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "raster = np.random.randint(0, 255, (1000, 1000))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Create the instance of triskele object with the array to filter."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "ap_random = triskele.Triskele(raster)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Filter the raster with this method :"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "filtered_raster = ap_random.filter()\n",
    "print(filtered_raster)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "If there is no error so far everything is okay.\n",
    "\n",
    "Theoretically the filtered output should be a copy of the input. (it's not, there is something going on along the edges, we should tell François !) \n",
    "\n",
    "## Attribute filter\n",
    "\n",
    "With `filter` method you can ask for specific attributes filtering. The results are stacked."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "filtered_rasters = ap_random.filter(area=[10, 100, 1e3])\n",
    "print(filtered_rasters.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "I don't know if you can see the signature with yout IDE : `filter(tree='max-tree', area=None, standard_deviation=None, moment_of_inertia=None)`\n",
    "\n",
    "The parameters are the same than apGenerator from François, I should write the doc..."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Advanced test\n",
    "\n",
    "With a real example.\n",
    "\n",
    "Load a TIFF file with 4 bands :"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "raster = triskele.read('../triskele/data/10m.tif')\n",
    "print(raster.shape)\n",
    "\n",
    "plt.imshow(raster)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Create the instance with some tweaks."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "ap_10m = triskele.Triskele(raster, verbose=False) # Verbose False to hide apGenerator gibbering :p"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Filter as you wish :"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "tree = 'tos-tree' # {'max-tree', 'min-tree', ...}\n",
    "area = [100, 1e3, 1e4]\n",
    "standard_deviation = None\n",
    "moment_of_inertia = None\n",
    "\n",
    "filtered_rasters = ap_10m.filter(tree=tree, area=area, standard_deviation=standard_deviation, moment_of_inertia=moment_of_inertia)\n",
    "print(filtered_rasters.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Display results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "for i in range(filtered_rasters.shape[2]):\n",
    "    plt.imshow(filtered_rasters[:,:,i])\n",
    "    plt.show()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}