environment.yml

name: hvplot_example
channels:
  - conda-forge
  - defaults
dependencies:
  - python=3.10
  - hvplot
  - intake
  - intake-parquet
  - intake-xarray
  - s3fs
  - datashader

hvplot_datashade_legend.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "e3a23a06-adc3-4a7a-a9f9-d52052fe6a2b",
   "metadata": {},
   "source": [
    "# Trying to get datashaded/rasterized hvPlots to show a legend\n",
    "\n",
    "This example uses an hvplot sample data set. As this sample dataset is actually quite small (~ 1 million records) datashading/rasterizing is not very beneficial. HOWEVER, in the case that you have 100 million or more datapoints, datashade/raster is required to generate a plot without crashing the browser (at least in my experience)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "aae7024d-701b-40cf-9d2c-a07284a9be6b",
   "metadata": {},
   "outputs": [],
   "source": [
    "# load hvplot and default bokeh extension\n",
    "import hvplot.pandas\n",
    "import holoviews as hv\n",
    "# load some sample data\n",
    "import hvplot.sample_data\n",
    "# pandas!\n",
    "import pandas as pd\n",
    "# colormaps\n",
    "import colorcet as cc"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "0e4ba4f1-0af8-4cc4-a607-ce89d4e29ac1",
   "metadata": {},
   "outputs": [],
   "source": [
    "df = hvplot.sample_data.airline_flights.read()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "bb5424d0-3bba-431b-8916-ca717ccea969",
   "metadata": {},
   "outputs": [],
   "source": [
    "df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b82adb1b-e8ae-4e20-9e60-e136a201d9c8",
   "metadata": {},
   "outputs": [],
   "source": [
    "df['dayofweek'] = df.dayofweek.astype('category')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ae8c032e-2cf5-4e00-b3a4-c029393b40f5",
   "metadata": {},
   "outputs": [],
   "source": [
    "df[df.carrier=='OO'].dayofweek.unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b56603a2-2969-468e-9240-120a789f249c",
   "metadata": {},
   "outputs": [],
   "source": [
    "df[df.carrier=='OO'].hvplot.scatter(\n",
    "    x='arr_time',\n",
    "    y='dep_time',\n",
    "    by='dayofweek'\n",
    ") + df[df.carrier=='OO'].hvplot.scatter(x='arr_time',\n",
    "                                        y='dep_time',\n",
    "                                        by='dayofweek',\n",
    "                                        datashade=True\n",
    "                                       )"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e3ee18c7-99b0-4efc-95fd-cc8946e8b2cd",
   "metadata": {},
   "source": [
    "## Here's a hack I figured out to have a legend and a datashaded plot"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "32263f26-10fd-4d0b-8d35-1c138a82b943",
   "metadata": {},
   "outputs": [],
   "source": [
    "arr = []\n",
    "for i in range(1, len(df.dayofweek.unique()) + 1):\n",
    "    arr.append(df[(df.carrier=='OO') & (df.dayofweek == i)].head(1))\n",
    "minidf = pd.concat(arr)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "59b31f83-748f-4e12-8ddb-fb7369615856",
   "metadata": {},
   "outputs": [],
   "source": [
    "minidf.hvplot.scatter(\n",
    "    x='arr_time',\n",
    "    y='dep_time',\n",
    "    by='dayofweek',\n",
    "    s=1\n",
    ") * df[df.carrier=='OO'].hvplot.scatter(x='arr_time',\n",
    "                                        y='dep_time',\n",
    "                                        by='dayofweek',\n",
    "                                        datashade=True\n",
    "                                       )"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "46e055ee-0fb9-484b-b7a9-ad4dd9908fc1",
   "metadata": {},
   "source": [
    "## Recommended to try rasterize to see if that would generate a legend"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b660bfb1-e630-4461-89b5-a711322e1088",
   "metadata": {},
   "outputs": [],
   "source": [
    "by_plot = None\n",
    "by_plot = df[df.carrier=='OO'].hvplot.scatter(x='arr_time',\n",
    "                                              y='dep_time',\n",
    "                                              by='dayofweek',\n",
    "                                              rasterize=True\n",
    "                                             )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b0b13940-0dac-4af0-b335-ab5c62b05219",
   "metadata": {},
   "outputs": [],
   "source": [
    "by_plot"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8a8a6b93-82a2-4e95-b883-8493bd883ce5",
   "metadata": {},
   "outputs": [],
   "source": [
    "print(by_plot)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "bdb799c7-6cb3-4a55-8de1-b7830fd40983",
   "metadata": {},
   "outputs": [],
   "source": [
    "type(by_plot)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "0466c9d4-3491-40c0-bec5-faa8115d5950",
   "metadata": {},
   "outputs": [],
   "source": [
    "hv.help(by_plot)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "1fb5f856-ddec-4955-8a8e-13c5e3ce8994",
   "metadata": {},
   "outputs": [],
   "source": [
    "by_plot.opts('Image', color_levels=7)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python [conda env:waveform] *",
   "language": "python",
   "name": "conda-env-waveform-py"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.11.6"
  },
  "nteract": {
   "version": "0.28.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}

hvplot_example.ipynb