test-notebook

test.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Slope Comparisons"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import altair as alt\n",
    "from scipy import stats\n",
    "from altair import datum\n",
    "from ipywidgets import interact\n",
    "\n",
    "# Set up some global config and variables\n",
    "alt.renderers.enable('default')\n",
    "pd.options.mode.chained_assignment = None\n",
    "np.seterr(all='ignore')\n",
    "\n",
    "df = pd.read_csv('https://raw.githubusercontent.com/ironhacks/COVID-19-notebook-demo-1/master/jhu-daily-reports.csv')\n",
    "df['Active']  = df.Confirmed - (df.Deaths + df.Recovered)\n",
    "samples = df[['Date', 'Country']].groupby('Date').Country.nunique()\n",
    "days = samples[samples > 1].index.tolist()\n",
    "df = df[df['Date'].isin(days)]\n",
    "\n",
    "country_level = df.groupby(['Country', 'Date'], as_index=False).sum()\n",
    "def state_data(country):\n",
    "    return df[df['Country'] == country].groupby(['State', 'Date'], as_index=False).sum()\n",
    "def county_data(state):\n",
    "    return df[(df['Country'] == 'US') & (df['State'] == state)].groupby(['County', 'Date'], as_index=False).sum()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def slope_chart(data, by, offset, xscale='linear', limit=400, scale=1, value='Confirmed_New', window=7, today=days[-1]):\n",
    "    data = data[data['Date']<=today]\n",
    "    source = data[data['Date'] == today]\n",
    "    for var in source[by].unique():\n",
    "        values = data[data[by] == var].sort_values('Date').tail(window)[['Confirmed', value]]\n",
    "        slope, intercept, r_value, p_value, std_err = stats.linregress(values.Confirmed, values[value])\n",
    "        source.loc[source[by] == var, 'Slope'] = slope\n",
    "    source.fillna(0, inplace=True)\n",
    "    source = source[source[value] > limit]\n",
    "\n",
    "    title = 'Slope of %s in last %d days since %s vs. Total Confirmed' % (' '.join(value.split('_')), window, today)\n",
    "    base = alt.Chart(source, title=title).mark_point(filled=True, stroke='grey').encode(\n",
    "        alt.X('Confirmed:Q', scale=alt.Scale(type=xscale), axis=alt.Axis(offset=offset)),\n",
    "        alt.Y('Slope:Q', axis=alt.Axis(title='Slope')),\n",
    "        color=alt.Color(by+':N', scale=alt.Scale(scheme='category20'), legend=alt.Legend(columns=2, clipHeight=20, padding=10)),\n",
    "        size=alt.Size(value+':Q', scale=alt.Scale(domain=[source.Confirmed_New.min(), source.Confirmed_New.max()], range=[100*scale, 3000*scale])),\n",
    "        tooltip=[by, 'Confirmed', 'Slope', value]\n",
    "    )\n",
    "    text = base.mark_text().encode(\n",
    "        text=by+':N',\n",
    "        size=alt.value(12),\n",
    "        color=alt.value('black')\n",
    "    ).transform_filter(datum[value] > limit*2)\n",
    "    regression = base.transform_regression('Confirmed', 'Slope', method=\"poly\", order=1).mark_line(strokeDash=[6,8]).encode(color=alt.value('grey'), size=alt.value(2))\n",
    "    hline = alt.Chart().mark_rule(color='red', strokeDash=[6,3]).encode(alt.Y('a:Q', axis=alt.Axis(title=''))).transform_calculate(a=\"0\")\n",
    "\n",
    "    return (base+text+regression+hline) if offset == 0 else (base+text+regression)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Country Level"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "slope_chart(country_level, 'Country', 0, xscale='log', limit=450, scale=3, window=7).properties(\n",
    "    width=1200,\n",
    "    height=800\n",
    ").interactive()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = country_level\n",
    "state = alt.Chart(data[data['Country'] == 'US'].sort_values('Date').tail(60)).mark_line().encode(\n",
    "    alt.X('Date:T', axis=alt.Axis(title='Cumulative Cases')),\n",
    "    alt.Y('Confirmed_New:Q', axis=alt.Axis(title='New Cases'))\n",
    ")\n",
    "reg = state.transform_regression(\"Date\", \"Confirmed_New\", method=\"linear\").mark_line(color='red', strokeDash=[6,3])\n",
    "(state+reg).interactive()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# State Level"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "@interact(window=(2, 21, 1))\n",
    "def chart(window=7):\n",
    "    return slope_chart(state_data('US'), 'State', 0, limit=100, xscale='log', scale=3, window=window).properties(\n",
    "        width=1200,\n",
    "        height=800\n",
    "    ).interactive()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = state_data('US')\n",
    "state = alt.Chart(data[data['State'] == 'OK'].sort_values('Date').tail(60)).mark_line().encode(\n",
    "    x='Confirmed:Q',\n",
    "    y='Confirmed_New:Q'\n",
    ")\n",
    "reg = state.transform_regression(\"Confirmed\", \"Confirmed_New\", method=\"poly\").mark_line(color='red', strokeDash=[6,3])\n",
    "(state+reg).interactive()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# US County Level"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "@interact(window=(2, 21, 1))\n",
    "def chart(window=7):\n",
    "    return slope_chart(county_data('CA'), 'County', 0, xscale='log', limit=15, scale=5, window=window).properties(\n",
    "        width=1100,\n",
    "        height=600\n",
    "    ).interactive()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "jupyter": {
     "source_hidden": true
    }
   },
   "outputs": [],
   "source": [
    "data = county_data('CA')\n",
    "state = alt.Chart(data[data['County'] == 'Los Angeles'].sort_values('Date').tail(60)).mark_line().encode(\n",
    "    x='Confirmed:Q',\n",
    "    y='Confirmed_New:Q'\n",
    ")\n",
    "reg = state.transform_regression(\"Confirmed\", \"Confirmed_New\", method=\"poly\").mark_line(color='red', strokeDash=[6,3])\n",
    "(state+reg).interactive()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# US Hospitalizations"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e741db7d90ba48129e8bc8f0ceeb2dee",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "interactive(children=(IntSlider(value=7, description='window', max=21, min=2), Output()), _dom_classes=('widge…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "dfh = pd.read_csv('https://covidtracking.com/api/v1/states/daily.csv')\n",
    "dfh.date = pd.to_datetime(dfh.date, format='%Y%m%d')\n",
    "dfh.date = dfh.date.dt.strftime('%m-%d-%Y')\n",
    "dfh = dfh.rename({'date': 'Date', 'state':'State', 'hospitalizedCurrently': 'Hospitalized'}, axis=1)\n",
    "data = state_data('US')\n",
    "data = data.merge(dfh, on=['Date', 'State'], how='outer')\n",
    "@interact(window=(2, 21, 1))\n",
    "def chart(window=7):\n",
    "    return slope_chart(\n",
    "        data, \n",
    "        'State', \n",
    "        0, \n",
    "        xscale='log', \n",
    "        limit=200, \n",
    "        scale=2, \n",
    "        value='Hospitalized', \n",
    "        window=window, \n",
    "        today=days[-1]\n",
    "    ).properties(\n",
    "        width=1100,\n",
    "        height=800\n",
    "    ).interactive()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "state = alt.Chart(data[data['State'] == 'CA'].sort_values('Date').tail(7)).mark_line().encode(\n",
    "    x='Date:T',\n",
    "    y='Hospitalized:Q'\n",
    ")\n",
    "reg = state.transform_regression(\"Date\", \"Hospitalized\", method=\"poly\").mark_line(color='red', strokeDash=[6,3])\n",
    "(state+reg).interactive()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = country_level\n",
    "state = alt.Chart(data[(data['Country'] == 'India')]).mark_line().encode(\n",
    "    alt.X('Date:T', axis=alt.Axis(title='Date')),\n",
    "    alt.Y('Confirmed_New:Q', axis=alt.Axis(title='New Cases'))\n",
    ")\n",
    "reg = state.transform_regression(\"Date\", \"Confirmed_New\", method=\"linear\").mark_line(color='grey', strokeDash=[6,3])\n",
    "\n",
    "marks = pd.DataFrame([\n",
    "    {\"Phase\": \"1\", \"start\": \"03-25-2020\", \"end\": \"04-14-2020\"},\n",
    "    {\"Phase\": \"2\", \"start\": \"04-14-2020\", \"end\": \"05-03-2020\"},\n",
    "    {\"Phase\": \"3\", \"start\": \"05-03-2020\", \"end\": \"05-17-2020\"},\n",
    "    {\"Phase\": \"4\", \"start\": \"05-17-2020\", \"end\": \"05-31-2020\"},\n",
    "])\n",
    "rect = alt.Chart(marks).mark_rect(opacity=0.3).encode(x='start:T', x2='end:T', color='Phase:N')\n",
    "\n",
    "(rect+state+reg).properties(\n",
    "    width=800,\n",
    "    height=500\n",
    ").interactive()"
   ]
  }
 ],
 "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",
   "version": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}