{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from lets_plot import *\n",
    "import lets_plot.mapping as pm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "LetsPlot.setup_html()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = {\n",
    "    'x': [0, 5, 10, 15],\n",
    "    'y': [0, 5, 10, 15],\n",
    "    'a': [1, 2, 3, 4],\n",
    "    'b': [4, 5, 6, 7]\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# as_discrete, no scale\n",
    "p = ggplot(df, aes(x='x', y='y')) + geom_point(aes(color=pm.as_discrete('a', label='custom name')), size=9)\n",
    "p"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# as_discrete, scale\n",
    "p = ggplot(df, aes(x='x', y='y')) \\\n",
    "    + geom_point(aes(color='a', fill=pm.as_discrete('b')), shape=21, size=9, stroke=5) \\\n",
    "    + scale_color_discrete()\n",
    "p"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# data in mappings, scale_color_discrete\n",
    "p = ggplot() + geom_point(aes(x=[0, 5, 10], y=[0, 5, 10], color=[1, 2, 4]), size=9) + scale_color_discrete()\n",
    "p"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Issue 1\n",
    "\n",
    "`as_discrete` is not working when used in \"ggplot()\" mapping."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "p3 = ggplot(df, aes(x='x', y='y', color=pm.as_discrete('a'))) + geom_point(size=9)\n",
    "p3"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Issue 2\n",
    "`as_discrete` doesn't create groups the way discrete variable does."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = {\n",
    "    'x': [0, 5, 10, 15],\n",
    "    'y': [0, 5, 10, 15],\n",
    "    'a': [0, 0, 1, 1],\n",
    "    'c': ['a', 'a', 'b', 'b']\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "p4 = ggplot(df, aes(x='x', y='y')) + geom_line(aes(color=pm.as_discrete('a')), size=3)\n",
    "p4"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# expected: 2 lines ('c' is a discrete variable)\n",
    "p5 = ggplot(df, aes(x='x', y='y')) + geom_line(aes(color='c'), size=3)\n",
    "p5"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# expected: 2 lines (`group` is defined manually)\n",
    "p6 = ggplot(df, aes(x='x', y='y')) + geom_line(aes(color='a', group='a'), size=3)\n",
    "p6"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Issue 2a\n",
    "Also about groups but with `stat` this time"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "mpg_df = pd.read_csv('https://raw.githubusercontent.com/JetBrains/lets-plot-docs/master/data/mpg.csv')\n",
    "mpg_plot = ggplot(mpg_df, aes(x='displ', y='hwy'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "cyl_factor = pm.as_discrete('cyl')\n",
    "mpg_plot + geom_point(aes(color=cyl_factor)) \\\n",
    "+ geom_smooth(aes(color=cyl_factor), method='lm', size=1, se=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# expected: seperate regression line for each group\n",
    "mpg_plot + geom_point(aes(color=cyl_factor)) \\\n",
    "+ geom_smooth(aes(color='cyl', group='cyl'), method='lm', size=1, se=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Issue 3\n",
    "\"Not an aesthetic 'group'\" error when used with `as_discrete`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "mpg_plot + geom_point(aes(color=cyl_factor)) \\\n",
    "+ geom_smooth(aes(color=cyl_factor, group='cyl'), method='lm', size=1, se=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Issue 4\n",
    "Nice to have parameter `ordered`. Owerwise have to use `scale_discrete(breaks=[...])` to order groups by number of cylinders:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "mpg_plot + geom_point(aes(color=cyl_factor)) \\\n",
    "+ scale_color_discrete(breaks=[4, 5, 6, 8])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "mpg_plot + geom_point(aes(color=pm.as_discrete('cyl', label='cyl')))"
   ]
  }
 ],
 "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": "3.7.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}