{ "cells": [ { "cell_type": "markdown", "metadata": {}, "source": [ "# Approximative Zwischenergebnisse des Nowcastings aus dem RKI Situationsbericht\n", "\n", "Thomas Viehmann, \n", "\n", "Hier extrahieren wir die Zahlen für Bekannte/Imputierte/Nowcast-Fälle aus den Situationsberichten.\n", "\n", "Bitte beachten Sie die Datenlizenz für die RKI Daten." ] }, { "cell_type": "code", "execution_count": 16, "metadata": {}, "outputs": [], "source": [ "import collections\n", "import bs4\n", "import pandas\n", "%matplotlib inline\n", "from matplotlib import pyplot\n", "import math\n", "import numpy\n", "import re\n", "import datetime" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Ausgangspunkt ist die SVG-Convertierte Seite mit der Nowcasting grafik aus dem Situationsbericht:\n", "\n", "https://www.rki.de/DE/Content/InfAZ/N/Neuartiges_Coronavirus/Situationsberichte/Gesamt.html\n", "\n", "Die kann man z.B. durch `pdf2svg 2020-05-16-de.pdf 2020-05-16-de.svg 7` oder mit Inkscape bekommen.\n", "Das ganze habe ich an einigen Tagen ausprobiert, es ist möglicherweise aber nicht stabil, weil es die \"unendlichen Möglichkeiten\" des SVG-Formats nicht abbildet.\n", "\n", "Wir benutzen die vom RKI veröffentlichten Nowcasting-Zahlen des RKI, um die Skalierung zu berechnen (eleganter, da eigenständiger wäre natürlich, eine Horizontale Achse (z.B. 4000) im Graphen zu suchen.\n", "\n", "https://www.rki.de/DE/Content/InfAZ/N/Neuartiges_Coronavirus/Projekte_RKI/Nowcasting.html\n", "\n", "Den Verlauf nehmen wir auch gleich zur Kontrolle.\n", "\n", "Hier sind die Dateinamen:" ] }, { "cell_type": "code", "execution_count": 226, "metadata": {}, "outputs": [], "source": [ "fn_nowcasting_excel = 'Nowcasting_Zahlen.2020-05-17.xlsx'\n", "fn_svg_from_report = '/tmp/2020-05-17-de.svg'\n", "fn_output = './nowcast_imputed_known_from_graph_2020-05-17.csv'" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Erst mal die Tabelle auslesen:" ] }, { "cell_type": "code", "execution_count": 311, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "5330" ] }, "execution_count": 311, "metadata": {}, "output_type": "execute_result" } ], "source": [ "rki_numbers = pandas.read_excel(fn_nowcasting_excel, sheet_name='Nowcast_R', index_col=0)\n", "nowcasting_excel = rki_numbers.iloc[:, 3]\n", "assert nowcasting_excel.name.startswith('Punkts')\n", "max_from_excel = nowcasting_excel.max()\n", "max_from_excel # should be around 5331, but does vary (presumably due to stochastic imputation)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Das extrahieren der Zahlen ist sehr ad-hoc. Wir finden mit den passenden Farben gefüllte `path`-Elemente und versuchen einen improvisierten Parser für die Bewegungen und merken uns die minimalen und maximalen x und y Koordinaten \"der Balken\"." ] }, { "cell_type": "code", "execution_count": 325, "metadata": {}, "outputs": [], "source": [ "root = bs4.BeautifulSoup(open(fn_svg_from_report))\n", "fill_re = re.compile('fill:rgb|fill:#')\n", "lookfor = {(21, 48, 64) : 'known', (56, 72, 85): 'nowcast', (70, 70, 70) : 'imputed',\n", " 'fill:#367ba4': 'known', 'fill:#b3b3b3': 'imputed', 'fill:#90bad9': 'nowcast'}\n", "paths_with_fill_style = [p for p in root.find_all('path') if fill_re.search(p.get('style',''))]\n", "\n", "values = collections.defaultdict(list)\n", "for p in paths_with_fill_style:\n", " fill_style = [s.strip() for s in p['style'].split(';') if fill_re.match(s) is not None][0]\n", " if 'rgb' in fill_style:\n", " rgb = tuple(int(float(s)) for s in re.match(r\"fill:rgb\\(([0-9\\.]+)%,([0-9\\.]+)%,([0-9\\.]+)%\\)\", fill_style).groups())\n", " else:\n", " rgb = fill_style.strip()\n", " key = lookfor.get(rgb)\n", " if key is not None:\n", " movements = re.split('[, ]+', p['d'].strip())\n", " i = 0\n", " ymin = math.inf\n", " ymax = -math.inf\n", " xmax = -math.inf\n", " xmin = math.inf\n", " while i < len(movements):\n", " v = movements[i]\n", " if v in {'M', 'L', 'm'}:\n", " x = float(movements[i + 1])\n", " y = float(movements[i + 2])\n", " i += 3\n", " elif v in {'Z', 'z'}:\n", " i += 1\n", " elif v == 'h':\n", " x += float(movements[i+1])\n", " i += 2\n", " elif v == 'H':\n", " x = float(movements[i+1])\n", " i += 2\n", " elif v == 'v':\n", " y += float(movements[i+1])\n", " i += 2\n", " elif v == 'V':\n", " y = float(movements[i+1])\n", " i += 2\n", " else:\n", " raise Exception(f\"need to parse more path: {i}, {v}\")\n", " xmin = min(xmin, x)\n", " ymin = min(ymin, y)\n", " ymax = max(ymax, y)\n", " xmax = max(xmax, x)\n", " values[key].append((xmin, ymin, ymax))" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Jertzt haben wir zu viel: mindestens mal aus der Legende bekommen wir \"Irrläufer\". Wir ermitteln eine \"Grundlinie\", die \"fast alle\" Pfade als Minimale/Maximale y-Ausdehnung haben. Ich kenne mich mit SVG nicht aus, aber ich habe den Eindruck, dass das Koordinatensystem sowohl unten als auch oben kleine y-Werte kann, deshalb checkt das Skript beides. Dann behalten wir nur Pfade, die auf der Grundline liegen." ] }, { "cell_type": "code", "execution_count": 326, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "nowcast 86 85\n", "imputed 86 85\n", "known 85 84\n" ] } ], "source": [ "values = dict(values)\n", "for k, v in values.items():\n", " vals = numpy.array(sorted(v))\n", " # drop legend\n", " p25 = numpy.percentile(vals, 25, axis=0)\n", " p75 = numpy.percentile(vals, 75, axis=0)\n", " iqr = p75-p25\n", " if iqr[1] < 1:\n", " values[k] = numpy.array([(x, ymax - ymin) for x, ymin, ymax in sorted(v) if numpy.abs(ymin-p75[1]) < 1])\n", " elif iqr[2] < 1:\n", " values[k] = numpy.array([(x, ymax - ymin) for x, ymin, ymax in sorted(v) if numpy.abs(ymax-p75[2]) < 1])\n", " else:\n", " raise Exception('Need xmin or xmax to vary little')\n", " print(k, len(vals), len(values[k]))\n" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Es scheint möglich, dass Werte (am Anfang?) fehlen. Also basteln wir Indizes aus der x-Position und füllen Leerstellen durch Nullen auf." ] }, { "cell_type": "code", "execution_count": 327, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "85" ] }, "execution_count": 327, "metadata": {}, "output_type": "execute_result" } ], "source": [ "xes = set()\n", "for v in values.values():\n", " xes |= {int(10*x) for x in v[:, 0]}\n", "xdict = {v: i for i,v in enumerate(sorted(xes))}\n", "\n", "for k, v in values.items():\n", " new_v = numpy.zeros(len(xdict))\n", " for x, h in v:\n", " new_v[xdict[int(10*x)]] = h\n", " values[k] = new_v\n", "len(xdict)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Jetzt skalieren wir alles mit dem Maximum und schon können wir graphen machen." ] }, { "cell_type": "code", "execution_count": 338, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "mult = max_from_excel / values['nowcast'].max()\n", "\n", "for k, v in values.items():\n", " v *= mult\n", " pyplot.plot(values[k], label=k)\n", "\n", "pyplot.legend()\n", "pyplot.savefig(fn_output+'.svg')\n", "pyplot.show()\n" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Dann machen wir einen Dataframe. Wir gehen davon aus, dass der Graph und die Nowcasting-Tabelle am selben Tag enden." ] }, { "cell_type": "code", "execution_count": 333, "metadata": {}, "outputs": [], "source": [ "end_date = nowcasting_excel.index.max()\n", "idx = pandas.date_range(end_date - pandas.Timedelta(len(values['nowcast'])-1, unit='D'), end_date, name='Refdatum')\n", "df = pandas.DataFrame(values, index=idx)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Vergleichen wir noch mit den Excel-Zahlen. Ich bekomme Abweichungen von 0-4." ] }, { "cell_type": "code", "execution_count": 334, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "[]" ] }, "execution_count": 334, "metadata": {}, "output_type": "execute_result" }, { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "pyplot.plot((nowcasting_excel - df.nowcast).dropna().abs())" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Wenn alles gut aussieht, sind wir fertig. Wir schreiben die Ergebnisse in ein CSV." ] }, { "cell_type": "code", "execution_count": 336, "metadata": {}, "outputs": [], "source": [ "df.to_csv(fn_output)" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [] } ], "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": "3.8.3rc1" } }, "nbformat": 4, "nbformat_minor": 2 }