altair-2023-10-22-AM4-thinning.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "d6c12234-a766-44c8-9fb6-aae38c1afdf2",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "┌──────────────┐\n",
       "│ count_star() │\n",
       "│    int64     │\n",
       "├──────────────┤\n",
       "│     50000000 │\n",
       "└──────────────┘"
      ]
     },
     "execution_count": 1,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import duckdb\n",
    "import altair as alt\n",
    "from ipywidgets import HTML, HBox\n",
    "\n",
    "np.random.seed(0)\n",
    "no_steps = 50000000\n",
    "steps = np.random.choice([-1, 1], no_steps)\n",
    "random_walk = np.cumsum(steps)\n",
    "\n",
    "df = pd.DataFrame(random_walk, columns=['y'])\n",
    "df.index.name = 'x'\n",
    "df = df.reset_index()\n",
    "\n",
    "con = duckdb.connect()\n",
    "con.register('df', df)\n",
    "\n",
    "con.query('SELECT count() FROM df')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "495de73b-3da1-4350-89c6-cc60a260fa7b",
   "metadata": {},
   "outputs": [],
   "source": [
    "def m4data(con, width, lowerbound, upperbound):\n",
    "    cur = con.cursor()\n",
    "    df_m4 = cur.query(f'''\n",
    "        SELECT \n",
    "            min(x) AS min_x, \n",
    "            arg_min(y, x) AS argmin_yx, \n",
    "            max(x) AS max_x, \n",
    "            arg_max(y, x) AS argmax_yx, \n",
    "            min(y) AS min_y, \n",
    "            arg_min(x, y) AS argmin_xy, \n",
    "            max(y) AS max_y, \n",
    "            arg_max(x, y) AS argmax_xy, \n",
    "            round({width} * (x - {lowerbound}) / ({upperbound} - {lowerbound})) AS bin\n",
    "        FROM df\n",
    "        WHERE x BETWEEN {lowerbound} AND {upperbound}\n",
    "        GROUP BY bin   \n",
    "    ''')    \n",
    "    \n",
    "    return df_m4.df()\n",
    "\n",
    "def chart_original(df):\n",
    "    x_range = [df.min_x.min(), df.max_x.max()]\n",
    "    interval = alt.selection_interval(encodings=['x'], bind='scales', name='interval', value={\"x\":x_range})\n",
    "    bind_range = alt.binding_range(min=100, max=400, name='Chart width: ')\n",
    "    param_width = alt.param('width', bind=bind_range, value=300)\n",
    "    \n",
    "    title=alt.Title(alt.expr(f'\"This chart is \" + {param_width.name} + \" px wide\"'))\n",
    "    chart = alt.Chart(df, title=title).mark_rect(tooltip=True).encode(\n",
    "        x='min_x:Q', \n",
    "        x2='max_x:Q', \n",
    "        y='min_y:Q', \n",
    "        y2='max_y:Q'\n",
    "    ).add_params(interval, param_width)\n",
    "    return chart\n",
    "\n",
    "def chart_updated(df, w):\n",
    "    chart = alt.Chart(df, width=w, title='auto chart').mark_line(\n",
    "        tooltip=True, \n",
    "        point=True,\n",
    "        interpolate='natural'\n",
    "    ).encode(\n",
    "        x=alt.X('min_x:Q').scale(zero=False, nice=False), \n",
    "        y=alt.Y('min_y:Q').scale(domain=[-6000, 12000])\n",
    "    )\n",
    "    return chart"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "79e9ab25-b64a-46ba-a1e3-bb42322e0555",
   "metadata": {},
   "outputs": [],
   "source": [
    "init_w = 300\n",
    "df_reduced = m4data(con, width=init_w, lowerbound=0, upperbound=no_steps)\n",
    "jchart_original = alt.JupyterChart(chart_original(df_reduced))\n",
    "jchart_updated = alt.JupyterChart(chart_updated(df_reduced, w=init_w))\n",
    "\n",
    "#HBox([jchart_original, jchart_updated])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "10641599-3af3-4251-a8ca-3b8eacefeda8",
   "metadata": {},
   "outputs": [],
   "source": [
    "# # manual tests\n",
    "# lb, ub = jchart_original.selections.interval.value['min_x']\n",
    "# w = jchart_original.params.width\n",
    "# print(w, lb, ub)\n",
    "# df_reduced_interactive = m4data(con, width=w, lowerbound=lb, upperbound=ub)\n",
    "# jchart_updated.chart = chart_updated(df_reduced_interactive, w)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "5fd8c29e-a552-49f7-8725-de2ac6c75389",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "4b40342f4b214bfd8cc7403fff53d324",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(JupyterChart(spec={'config': {'view': {'continuousWidth': 300, 'continuousHeight': 300}}, 'data…"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def on_width_change(width_change):\n",
    "    w = width_change.new\n",
    "    # w = jchart_original.params.width    \n",
    "    lb, ub = jchart_original.selections.interval.value['min_x']\n",
    "    \n",
    "    df_reduced_width = m4data(con, width=w, lowerbound=int(lb), upperbound=int(ub))\n",
    "    print(w, lb, ub, df_reduced_width.shape)\n",
    "    jchart_updated.chart = chart_updated(df_reduced_width, w=w)\n",
    "\n",
    "def on_interval_change(interval_change):\n",
    "    lb, ub = interval_change.new.value['min_x']\n",
    "    # lb, ub = jchart_original.selections.interval.value['min_x']\n",
    "    w = jchart_original.params.width\n",
    "\n",
    "    df_reduced_interval = m4data(con, width=w, lowerbound=int(lb), upperbound=int(ub))\n",
    "    print(w, lb, ub, df_reduced_interval.shape)\n",
    "    jchart_updated.chart = chart_updated(df_reduced_interval, w=w)\n",
    "\n",
    "jchart_original.params.observe(on_width_change, [\"width\"])\n",
    "jchart_original.selections.observe(on_interval_change, [\"interval\"])\n",
    "\n",
    "HBox([jchart_original, jchart_updated])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "52c18bd0-bd23-45e1-a7ce-653d092e045e",
   "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": "3.11.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}

Awesome gist