{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Tutorial about transforming LocData"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Locan provides methods for transforming localization data sets into new LocData objects."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from pathlib import Path\n",
    "\n",
    "%matplotlib inline\n",
    "\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "from mpl_toolkits.mplot3d import Axes3D\n",
    "\n",
    "import locan as lc"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "lc.show_versions(system=False, dependencies=False, verbose=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# A path in which test data can be found:\n",
    "TEST_DIR: Path = Path.cwd().parents[2] / \"tests\"\n",
    "TEST_DIR"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Spatially randomize a structured set of localizations"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Assume that localizations are somehow structured throughout a region. Often it is helpful to compare analysis results to a similar dataset in which localizations are homogeneously Poisson distributed. A LocData object with this kind of data can be provided by the randomize function."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "rng = np.random.default_rng(seed=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "outputs": [],
   "source": [
    "locdata = lc.simulate_Thomas(parent_intensity=1e-5, region=((0, 1000), (0, 1000)), cluster_mu=100, cluster_std=10, seed=rng)\n",
    "locdata.print_summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "locdata_random = lc.randomize(locdata, hull_region='bb', seed=rng)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "fig, ax = plt.subplots(nrows=1, ncols=2)\n",
    "locdata.data.plot.scatter(x='position_x', y='position_y', ax=ax[0], color='Blue', label='locdata')\n",
    "locdata_random.data.plot.scatter(x='position_x', y='position_y', ax=ax[1], color='Blue', label='locdata')\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "print('Area of bounding box for structured data: {:.0f}'.format(locdata.properties['region_measure_bb']))\n",
    "print('Area of bounding box for randomized data: {:.0f}'.format(locdata_random.properties['region_measure_bb']))\n",
    "print('Ratio: {:.4f}'.format(locdata_random.properties['region_measure_bb'] / locdata.properties['region_measure_bb']))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Regions other from bounding box can be specified as RoiRegion instance."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "region = lc.ConvexHull(locdata.coordinates).region\n",
    "locdata_random = lc.randomize(locdata, hull_region=region)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "fig, ax = plt.subplots(nrows=1, ncols=2)\n",
    "locdata.data.plot.scatter(x='position_x', y='position_y', ax=ax[0], color='Blue', label='locdata')\n",
    "locdata_random.data.plot.scatter(x='position_x', y='position_y', ax=ax[1], color='Blue', label='locdata')\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "print('Area of bounding box for structured data: {:.0f}'.format(locdata.properties['region_measure_bb']))\n",
    "print('Area of bounding box for randomized data: {:.0f}'.format(locdata_random.properties['region_measure_bb']))\n",
    "print('Ratio: {:.4f}'.format(locdata_random.properties['region_measure_bb'] / locdata.properties['region_measure_bb']))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Apply an affine transformation to localization coordinates"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "A wrapper function provides affine transformations based on either numpy or open3d methods."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "matrix = ((-1, 0), (0, -1))\n",
    "offset = (10, 10)\n",
    "pre_translation = (100, 100)\n",
    "\n",
    "locdata_transformed = lc.transform_affine(locdata, matrix, offset, pre_translation, method='numpy')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "fig, ax = plt.subplots()\n",
    "locdata.data.plot.scatter(x='position_x', y='position_y',color='Blue', label='locdata', ax=ax)\n",
    "locdata_transformed.data.plot.scatter(x='position_x', y='position_y', color='Red', label='locdata', ax=ax);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Apply a BunwarpJ transformation to localization coordinates"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Often a transformation matrix was computed using ImageJ. The `bunwarp` function allows applying a transformation from the raw matrix of the ImageJ/Fiji plugin BunwarpJ. Here we show a very small region with a single fluorescent bead that is recorded on a red and a green dSTORM channel."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "matrix_path = TEST_DIR / 'test_data/transform/BunwarpJ_transformation_raw_green.txt'\n",
    "locdata_green = lc.load_asdf_file(path=TEST_DIR /\n",
    "                                 'test_data/transform/rapidSTORM_beads_green.asdf')\n",
    "locdata_red = lc.load_asdf_file(path=TEST_DIR /\n",
    "                                 'test_data/transform/rapidSTORM_beads_red.asdf')\n",
    "\n",
    "locdata_green_transformed = lc.bunwarp(locdata=locdata_green, matrix_path=matrix_path, pixel_size=(10, 10), flip=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "fig, ax = plt.subplots()\n",
    "locdata_red.data.plot.scatter(x='position_x', y='position_y',color='Red', label='locdata_red', alpha=0.5, ax=ax)\n",
    "locdata_green_transformed.data.plot.scatter(x='position_x', y='position_y', color='Green', label='locdata_green_transformed', alpha=0.5, ax=ax)\n",
    "locdata_green.data.plot.scatter(x='position_x', y='position_y',color='Blue', label='locdata_green', alpha=0.5, ax=ax);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "fig, ax = plt.subplots(figsize=(10, 8))\n",
    "lc. render_2d_rgb_mpl([locdata_red, locdata_green_transformed, locdata_green], bin_size=5, bin_range=((200, 800), (700, 1400)), rescale=lc.Trafo.EQUALIZE_0P3, ax=ax);"
   ]
  }
 ],
 "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": "3.12.6"
  },
  "widgets": {
   "application/vnd.jupyter.widget-state+json": {
    "state": {},
    "version_major": 2,
    "version_minor": 0
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}