Upload 5 files

app.py

@@ -0,0 +1,35 @@
+import gradio as gr
+from utils import get_calculated_df, get_info, text2oddslist
+
+
+def calculate(amount, text):
+    df = get_calculated_df(amount, text2oddslist(text))
+    info = get_info(df)
+
+    # dfの整形
+    display_cols = ["odds", "buy", "refound"]
+    if df["name"].map(lambda name: name != "").any():
+        display_cols = ["name"] + display_cols
+
+    # markdownの整形
+    md_string = (
+        f"**購入額:** ¥ **{info['sum']:,}**<br>"
+        f"**点数:** **{info['num_kind']}** 点<br>"
+        f"**払戻:** ¥ **{info['refound_mean']:,}** (**{info['profit_mean']:+,}** (**{info['rate_min']:+.0%}**))<br>"
+        "---"
+    )
+
+    return md_string, df[["index"] + display_cols]
+
+
+amount_input = gr.Number(value=3000, label="amount")
+text_input = gr.Textbox(label="text")
+markdown_output = gr.Markdown()
+df_output = gr.Dataframe(type="pandas")
+iface = gr.Interface(
+    fn=calculate,
+    inputs=[amount_input, text_input],
+    outputs=[markdown_output, df_output],
+)
+
+iface.launch()

app_gradio.py

@@ -0,0 +1,35 @@
+import gradio as gr
+from utils import get_calculated_df, get_info, text2oddslist
+
+
+def calculate(amount, text):
+    df = get_calculated_df(amount, text2oddslist(text))
+    info = get_info(df)
+
+    # dfの整形
+    display_cols = ["odds", "buy", "refound"]
+    if df["name"].map(lambda name: name != "").any():
+        display_cols = ["name"] + display_cols
+
+    # markdownの整形
+    md_string = (
+        f"**購入額:** ¥ **{info['sum']:,}**<br>"
+        f"**点数:** **{info['num_kind']}** 点<br>"
+        f"**払戻:** ¥ **{info['refound_mean']:,}** (**{info['profit_mean']:+,}** (**{info['rate_min']:+.0%}**))<br>"
+        "---"
+    )
+
+    return md_string, df[["index"] + display_cols]
+
+
+amount_input = gr.Number(value=3000, label="amount")
+text_input = gr.Textbox(label="text")
+markdown_output = gr.Markdown()
+df_output = gr.Dataframe(type="pandas")
+iface = gr.Interface(
+    fn=calculate,
+    inputs=[amount_input, text_input],
+    outputs=[markdown_output, df_output],
+)
+
+iface.launch()

app_streamlit.py

@@ -0,0 +1,27 @@
+import streamlit as st
+from utils import get_calculated_df, get_info, text2oddslist
+
+amount: int = int(st.number_input("amount", min_value=100, value=3000, step=100))
+text: str = st.text_area("text", "")
+# ボタンを作る
+st.button("calculate")
+if text:
+    df = get_calculated_df(amount, text2oddslist(text))
+    # dataframeを綺麗に出力する
+    info = get_info(df)
+    col1, col2 = st.columns(2)
+    with col1:
+        st.markdown(f"#### 購入額: ¥ **{info['sum']:,}**")
+        st.markdown(f"#### 点　数: **{info['num_kind']}** 点")
+    with col2:
+        st.metric(
+            label="払戻",
+            value=f"¥ {info['refound_mean']:,}",
+            delta=f"{info['profit_mean']:+,}({info['rate_min']:.0%})",
+        )
+    st.markdown("---")
+    # dfの表示
+    display_cols = ["odds", "buy", "refound"]
+    if df["name"].map(lambda name: name != "").any():
+        display_cols = ["name"] + display_cols
+    st.dataframe(df[display_cols])

requirements.txt

@@ -0,0 +1,3 @@
+pandas==2.1.0
+streamlit==1.26.0
+gradio==3.44.3

utils.py

@@ -0,0 +1,107 @@
+import math
+from typing import TypedDict
+
+import pandas as pd
+
+
+# Typing -----------------------------------------------------------------------
+class OddsInfo(TypedDict):
+    index: str
+    name: str
+    odds: float
+
+
+OdddsList = list[OddsInfo]
+
+
+class Info(TypedDict):
+    sum: int
+    num_kind: int
+    refound_mean: int
+    profit_mean: int
+    rate_min: float
+    rate_max: float
+    rate_mean: float
+
+
+# text　->　odds_list ------------------------------------------------------------
+def _text2oddslist_pc(text: str) -> OdddsList:
+    odds_list: OdddsList = []
+    tmp_list: list[str] = []
+    lines: list[str] = text.split("\n")
+    for text in lines:
+        if "." in text:
+            odds_list.append(
+                {
+                    "index": "-".join(tmp_list),
+                    "name": "",
+                    "odds": float(text),
+                }
+            )
+            tmp_list = []
+        else:
+            if text.strip():
+                tmp_list.append(text.strip())
+    return odds_list
+
+
+def _text2oddslist_mobile(text: str) -> OdddsList:
+    odds_list: OdddsList = []
+    lines: list[str] = text.split("\n")
+    if "選択" in lines:
+        lines = lines[lines.index("選択") + 1 :]
+    if "すべてを選択" in lines:
+        lines = lines[: lines.index("すべてを選択")]
+    for i in range(len(lines) // 3):
+        # oddsが1つ目にある
+        if "." in lines[3 * i + 1]:
+            idx_odds = 1
+            idx_name = 2
+        else:
+            idx_odds = 2
+            idx_name = 1
+
+        odds_list.append(
+            {
+                "index": "-".join(lines[3 * i].split()),
+                "name": "-".join(lines[3 * i + idx_name].split()),
+                "odds": float(lines[3 * i + idx_odds]),
+            }
+        )
+    return odds_list
+
+
+def text2oddslist(text: str) -> OdddsList:
+    # pcの時
+    if set("".join(text.split())) <= set("0123456789.-"):
+        odds_list = _text2oddslist_pc(text)
+    # moduleの時
+    else:
+        odds_list = _text2oddslist_mobile(text)
+    return odds_list
+
+
+# odds_list -> df ---------------------------------------------------------------
+def get_calculated_df(amount: int, odds_list: OdddsList) -> pd.DataFrame:
+    df = pd.DataFrame(odds_list)
+    df["weight"] = (1 / df["odds"]) / (1 / df["odds"]).sum()
+    df["buy"] = (df["weight"] * amount / 100).map(
+        lambda x: max(math.floor(x) * 100, 100)
+    )
+    df["refound"] = (df["buy"] * df["odds"]).astype(int)
+    df["profit"] = df["refound"] - amount
+    df = df
+    return df
+
+
+def get_info(df: pd.DataFrame) -> Info:
+    info: Info = {
+        "sum": df["buy"].sum(),
+        "num_kind": df.shape[0],
+        "refound_mean": int(df["refound"].mean()),
+        "profit_mean": int(df["profit"].mean()),
+        "rate_min": df["refound"].min() / df["buy"].sum(),
+        "rate_max": df["refound"].max() / df["buy"].sum(),
+        "rate_mean": df["refound"].mean() / df["buy"].sum(),
+    }
+    return info