Switch from Jupyter to isolated Marimo notebooksHEADmain

1 files changed, 286 insertions, 0 deletions

diff --git a/notebooks/2024-11-29-election-petition.py b/notebooks/2024-11-29-election-petition.py
new file mode 100644
index 0000000..0666c7f
--- /dev/null
+++ b/notebooks/2024-11-29-election-petition.py
@@ -0,0 +1,286 @@
+# /// script
+# requires-python = ">=3.13"
+# dependencies = [
+#     "marimo",
+#     "matplotlib==3.10.5",
+#     "pandas==2.3.1",
+#     "python-dotenv==1.1.1",
+#     "seaborn==0.13.2",
+# ]
+# ///
+
+import marimo
+
+__generated_with = "0.14.17"
+app = marimo.App()
+
+
+@app.cell(hide_code=True)
+def _(mo):
+    mo.md(
+        r"""
+    # Analysing the Petition for a General Election
+
+    2024 general election results where downloaded from [here](https://commonslibrary.parliament.uk/research-briefings/cbp-10009/) on 29 November 2024. Petition results were downloaded using the following command on 29 November 2024:
+
+    ```shell
+    curl https://petition.parliament.uk/petitions/700143.json | jq > data/election-petition-results.json
+    ```
+    """
+    )
+    return
+
+
+@app.cell
+def _():
+    import json
+    from pathlib import Path
+
+    import marimo as mo
+    import matplotlib.pyplot as plt
+    import pandas as pd
+    import seaborn as sns
+    from dotenv import find_dotenv
+
+    ROOT_DIR = Path(find_dotenv("pyproject.toml")).parent
+    DATA_DIR = ROOT_DIR / "data"
+    ELECTION_RESULTS_PATH = DATA_DIR / "election-petition" / "HoC-GE2024-results-by-candidate.csv"
+    PETITION_RESULTS_PATH = DATA_DIR / "election-petition" / "election-petition-results.json"
+
+    TOP_PARTIES = [
+        "Conservative",
+        "Labour",
+        "Liberal Democrat",
+        "Reform UK",
+        "Green",
+        "Scottish National Party",
+        "Plaid Cymru",
+        "Independent",
+    ]
+
+    PARTY_COLOURS = {
+        "Conservative": "#0087DC",
+        "Labour": "#E4003B",
+        "Liberal Democrat": "#FAA61A",
+        "Reform UK": "#12B6CF",
+        "Green": "#02A95B",
+        "Scottish National Party": "#FDF38E",
+        "Plaid Cymru": "#005B54",
+        "Independent": "grey",
+    }
+
+    sns.set_style("ticks")
+    sns.set_palette(PARTY_COLOURS.values())
+    return (
+        ELECTION_RESULTS_PATH,
+        PETITION_RESULTS_PATH,
+        TOP_PARTIES,
+        json,
+        mo,
+        pd,
+        sns,
+    )
+
+
+@app.cell
+def _(ELECTION_RESULTS_PATH, PETITION_RESULTS_PATH, json, pd):
+    with PETITION_RESULTS_PATH.open("r", encoding="utf-8") as f:
+        petition_results_json = json.load(f)
+
+    petition_results_df = (
+        pd.DataFrame.from_records(
+            petition_results_json["data"]["attributes"]["signatures_by_constituency"],
+            exclude=["name", "mp"],
+        )
+        .rename(columns={"ons_code": "ons_id"})
+    )
+
+    df = (
+        pd.read_csv(
+            ELECTION_RESULTS_PATH,
+            header=0,
+            usecols=[
+                "ONS ID",
+                "Constituency name",
+                "Region name",
+                "Party name",
+                "Votes",
+                "Share",
+            ]
+        )
+        .rename(columns={
+            "ONS ID": "ons_id",
+            "Constituency name": "constituency",
+            "Region name": "region",
+            "Party name": "party",
+            "Votes": "votes",
+            "Share": "share",
+        })
+        .replace({"party": {"Labour and Co-operative": "Labour"}})  # Labour in Wales
+        .merge(petition_results_df, on="ons_id", how="left")
+    )
+
+    df = (
+        df
+        .join(df.groupby("ons_id")["votes"].sum().rename("total_votes"), how="left", on="ons_id")
+        .assign(signatures_per_vote=lambda x: x["signature_count"] / x["total_votes"])
+    )
+
+    df = (
+        df.merge(
+            df
+            .sort_values("votes", ascending=False)
+            .groupby("ons_id")
+            .first()[["party", "votes"]]
+            .assign(winner=True)
+            .reset_index(),
+            on=["ons_id", "party", "votes"],
+            how="left",
+        )
+        .fillna({"winner": False})
+    )
+
+    print(f'Total petition signatures: {petition_results_json["data"]["attributes"]["signature_count"]:,d}')
+    print(f'Total petition signatures from per-consituency counts: {df.groupby("ons_id")["signature_count"].max().sum():,d}')
+    return (df,)
+
+
+@app.cell
+def _(df):
+    df
+    return
+
+
+@app.cell
+def _(df, sns):
+    ax = sns.histplot(
+        df.groupby("ons_id")["signature_count"].max()
+    )
+    ax.set(xlabel="Signatures", ylabel="Constituencies", title="Distribution of Petition Signatures by Constituency")
+    ax
+    return
+
+
+@app.cell
+def _(mo):
+    mo.md(r"""### Signature Distribution""")
+    return
+
+
+@app.cell
+def _(TOP_PARTIES, df, sns):
+    _top_n = 3
+    _g = sns.displot(
+        df[df['winner'] & df['party'].isin(TOP_PARTIES[:_top_n])],
+        kind='hist',
+        x='signature_count',
+        hue='party',
+        hue_order=TOP_PARTIES[:_top_n],
+        col='party',
+        col_order=TOP_PARTIES[:_top_n],
+        height=3,
+        legend=False,
+    ).set_titles('{col_name}').set_axis_labels('Signatures', 'Constituencies')
+    _g.figure
+    return
+
+
+@app.cell
+def _(mo):
+    mo.md(r"""### Signatures per 2024 General Election Vote""")
+    return
+
+
+@app.cell
+def _(TOP_PARTIES, df, sns):
+    _top_n = 3
+    _g = (
+        sns.displot(
+            df[df["winner"] & df["party"].isin(TOP_PARTIES[:_top_n])],
+            kind="hist",
+            x="signatures_per_vote",
+            hue="party",
+            hue_order=TOP_PARTIES[:_top_n],
+            col="party",
+            col_order=TOP_PARTIES[:_top_n],
+            height=3,
+            legend=False,
+        )
+        .set_titles("{col_name}")
+        .set_axis_labels("Signatures per Vote", "Constituencies")
+    )
+    _g.figure
+    return
+
+
+@app.cell
+def _(mo):
+    mo.md(r"""### Signatures vs. Votes""")
+    return
+
+
+@app.cell
+def _(TOP_PARTIES, df, sns):
+    _g = sns.lmplot(
+        df[df["party"].isin(TOP_PARTIES)],
+        x="votes",
+        y="signature_count",
+        col="party",
+        col_wrap=4,
+        col_order=TOP_PARTIES,
+        hue="party",
+        hue_order=TOP_PARTIES,
+        height=3,
+    ).set_titles("{col_name}").set_axis_labels("2024 General Election Votes", "Petition Signatures")
+    _g.figure
+    return
+
+
+@app.cell
+def _(mo):
+    mo.md(r"""### Signatures vs. Vote Share""")
+    return
+
+
+@app.cell
+def _(TOP_PARTIES, df, sns):
+    _g = sns.lmplot(
+        df[df["party"].isin(TOP_PARTIES)],
+        x="share",
+        y="signature_count",
+        col="party",
+        col_wrap=4,
+        col_order=TOP_PARTIES,
+        hue="party",
+        hue_order=TOP_PARTIES,
+        height=3,
+    ).set_titles("{col_name}").set_axis_labels("2024 General Election Vote Share", "Petition Signatures")
+    _g.figure
+    return
+
+
+@app.cell
+def _(mo):
+    mo.md(r"""### Signatures per Vote vs. Vote Share""")
+    return
+
+
+@app.cell
+def _(TOP_PARTIES, df, sns):
+    _g = sns.lmplot(
+        df[df["party"].isin(TOP_PARTIES)],
+        x="share",
+        y="signatures_per_vote",
+        col="party",
+        col_wrap=4,
+        col_order=TOP_PARTIES,
+        hue="party",
+        hue_order=TOP_PARTIES,
+        height=3,
+    ).set_titles("{col_name}").set_axis_labels("2024 General Election Vote Share", "Petition Signatures per Vote")
+    _g.figure
+    return
+
+
+if __name__ == "__main__":
+    app.run()