netkeibaのレース結果から隊列をplotlyで可視化

netkeibaのレース結果から隊列をplotlyで可視化.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "authorship_tag": "ABX9TyOzfWbYtiFz6AKeeY+jcW/5",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/gist/kun432/d29085ce93a0b4a2eb952dea626d172a/.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "zennの記事を元に以下を追加してみた。\n",
        "\n",
        "- netkeibaのレース結果をスクレイプして隊列情報を取得\n",
        "- 各コーナーでの内外の位置取りを追加\n",
        "- pandas-1.4.0でpd.appendが廃止されることに伴い、pd.concatに書き換え\n",
        "- plotlyで隊列を可視化\n",
        "\n",
        "参考）pyparsingで競馬のコーナー通過順位をパース\n",
        "https://zenn.dev/moripon/articles/ed5caa9c1d621e\n"
      ],
      "metadata": {
        "id": "qngtoogbjC2m"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "!pip install pandas\n",
        "!pip install plotly"
      ],
      "metadata": {
        "id": "46uGuY2ObEWu"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "UCipf59VE4EB"
      },
      "outputs": [],
      "source": [
        "import numpy as np\n",
        "import pandas as pd\n",
        "import pyparsing as pp\n",
        "import plotly.graph_objects as go\n",
        "from plotly.subplots import make_subplots\n",
        "import math\n",
        "\n",
        "# DataFrame列名定義\n",
        "columns = ['diff', 'horse_no', 'side']\n",
        "\n",
        "# 差の定数(unit:馬身)\n",
        "DIFF_GROUP = 0.3\n",
        "DIFF_MIN = 1.5\n",
        "DIFF_MID = 3.0\n",
        "DIFF_MUCH = 6.0\n",
        "\n",
        "class ParsePass():\n",
        "    \n",
        "    def __init__(self):\n",
        "        \n",
        "        # 馬番\n",
        "        horse_no = pp.Word(pp.nums).setParseAction(self._horse_no_action)\n",
        "        \n",
        "        # 馬群\n",
        "        group = pp.Suppress(pp.Literal('(')) + \\\n",
        "                    pp.Optional(pp.delimitedList(pp.Word(pp.nums), delim=',')) + \\\n",
        "                    pp.Suppress(pp.Literal(')'))\n",
        "        group.ignore('*')\n",
        "        group.setParseAction(self._group_action)\n",
        "\n",
        "        # 情報要素\n",
        "        element = (group | horse_no)\n",
        "        \n",
        "        # 前走馬との差\n",
        "        diff_min = pp.Suppress(pp.Optional(pp.Literal(','))).setParseAction(self._diff_min_action) + element\n",
        "        diff_mid = pp.Suppress(pp.Literal('-')).setParseAction(self._diff_mid_action) + element\n",
        "        diff_much = pp.Suppress(pp.Literal('=')).setParseAction(self._diff_much_action) + element\n",
        "\n",
        "        # 全体定義\n",
        "        self._passing_order = element + pp.ZeroOrMore( diff_mid | diff_much | diff_min )\n",
        "        \n",
        "    def _horse_no_action(self, token):\n",
        "        \n",
        "        df_append = pd.DataFrame(data=[[self._diff, token[0], 1]], columns=columns)\n",
        "        self._data = pd.concat([self._data, df_append], ignore_index=True, axis=0).drop_duplicates().reset_index(drop=True)\n",
        "        return\n",
        "\n",
        "    def _group_action(self, token):\n",
        "        \n",
        "        for i, no in enumerate(token):\n",
        "            df_append = pd.DataFrame(data=[[self._diff, no, 1+i]], columns=columns)\n",
        "            self._data = pd.concat([self._data, df_append], ignore_index=True, axis=0).drop_duplicates().reset_index(drop=True)\n",
        "            self._diff += DIFF_GROUP\n",
        "        self._diff -= DIFF_GROUP\n",
        "        return\n",
        "        \n",
        "    def _diff_min_action(self, token):\n",
        "        \n",
        "        self._diff += DIFF_MIN\n",
        "        return\n",
        "        \n",
        "    def _diff_mid_action(self, token):\n",
        "        \n",
        "        self._diff += DIFF_MID\n",
        "        return\n",
        "    \n",
        "    def _diff_much_action(self, token):\n",
        "        \n",
        "        self._diff += DIFF_MUCH\n",
        "        return\n",
        "        \n",
        "    def parse(self, pass_str):\n",
        "        \n",
        "        # 初期化\n",
        "        self._data = pd.DataFrame(columns=columns)\n",
        "        self._diff = 0\n",
        "        # parse\n",
        "        self._passing_order.parseString(pass_str)\n",
        "        # index調整\n",
        "        self._data.index = np.arange(1, len(self._data)+1)\n",
        "        self._data.index.name = 'rank'\n",
        "\n",
        "        return(self._data)"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "netkeibaのレース結果のURLを右下のフォームに入力\n",
        "\n",
        "※db.netkeiba.comではなくてrace.netkeba.comのほう\n"
      ],
      "metadata": {
        "id": "qoTCm3QDVZa5"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "url = \"https://race.netkeiba.com/race/result.html?race_id=202305010811\"#@param {type: \"string\"}"
      ],
      "metadata": {
        "id": "KndSiMxMkMCh"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "df_race = pd.read_html(url)[0]\n",
        "df_race.drop(columns=[\"性齢\",\"斤量\",\"騎手\",\"タイム\",\"着差\",\"馬体重 (増減)\",\"厩舎\"], inplace=True)\n",
        "\n",
        "df_corner = pd.read_html(url)[3]\n",
        "df_corner.rename(columns={0: \"corner\", 1: \"order\"}, inplace=True)\n",
        "df_corner.dropna(inplace=True)\n",
        "df_corner.reset_index(drop=True, inplace=True)\n",
        "\n",
        "pass_parsing = ParsePass()\n",
        "\n",
        "dfs = {}\n",
        "for row in df_corner.itertuples():\n",
        "  df = pass_parsing.parse(row.order)\n",
        "  df[\"horse_no\"] = df[\"horse_no\"].astype(int)\n",
        "  df_merged = pd.merge(df, df_race, left_on='horse_no', right_on=\"馬 番\")\n",
        "  dfs[row.corner] = df_merged\n",
        "\n",
        "color_map = {1:\"snow\", 2:\"black\", 3:\"red\", 4:\"blue\",5:\"yellow\",6:\"green\",7:\"orange\",8:\"pink\"}\n",
        "\n",
        "fig = make_subplots(rows=1, cols=len(dfs), subplot_titles=list(dfs.keys()))\n",
        "\n",
        "diff_max=0\n",
        "for i, k in enumerate(dfs):\n",
        "  df = dfs[k]\n",
        "  if df[\"diff\"].max() > diff_max:\n",
        "    diff_max=df[\"diff\"].max()\n",
        "\n",
        "diff_max = math.ceil(diff_max)\n",
        "print(diff_max)\n",
        "\n",
        "for i, k in enumerate(dfs):\n",
        "  df = dfs[k]\n",
        "  colors = [color_map[val] for val in df[\"枠\"]]\n",
        "  fig.add_trace(\n",
        "    go.Scatter(x=df[\"side\"], y=df[\"diff\"],\n",
        "                  mode = 'markers',\n",
        "                  marker = dict(size=25, color=colors, line_width=3),\n",
        "                  text=df[\"馬名\"],\n",
        "                  showlegend=False,\n",
        "                  hovertemplate=\"%{text} <br>\",\n",
        "\n",
        "     ), row=1, col=i+1\n",
        "  )\n",
        "  fig.update_xaxes(title = '外 ←←←←←  内外  →→→→→ 内', range=[6,0], dtick=1, visible=True, showticklabels=True, zerolinecolor='LimeGreen', gridcolor='LimeGreen' ,row=1,col=i+1)\n",
        "  fig.update_yaxes(title = '後　→→→→→　前後（馬身） →→→→→　前', range=[diff_max + 1, -1], zerolinecolor='LightGreen', gridcolor='LightGreen', dtick=5,showticklabels=True, row=1,col=i+1)\n",
        "\n",
        "fig.update_layout(\n",
        "    title = 'コーナー通過順',\n",
        "    width=len(dfs) * 300, height=diff_max * 40,\n",
        "    plot_bgcolor=\"LimeGreen\",\n",
        ")\n",
        "\n",
        "print(df_corner)\n",
        "fig.show()"
      ],
      "metadata": {
        "id": "yz4pqvI2Urau"
      },
      "execution_count": null,
      "outputs": []
    }
  ]
}

コメント失礼します。
素人質問で恐縮なのですが、私がこのコードを実行すると"馬体重"や"馬番"の部分でエラーが出てしまうのですがkun432様の環境ではエラーなしで実行できているのでしょうか？

@shaketora ちょっとadhocですが修正しました。ご指摘ありがとうございます！