app.py

import marimo

__generated_with = "0.13.0"
app = marimo.App(width="full")


@app.cell
def _():
    import marimo as mo
    import pandas as pd
    import plotly.graph_objects as go
    from sklearn.preprocessing import MinMaxScaler
    import matplotlib.cm as cm
    import matplotlib.colors as mcolors
    return cm, go, mcolors, mo, pd, MinMaxScaler


@app.cell
def _(MinMaxScaler, pd):
    # --- Load data
    df = pd.read_excel("plant_species_richness.xlsx")

    # Columns used in radar
    radar_cols = [
        "Woody_Spps",
        "Veg_Spps",
        "Bryo_Spps",
        "Coffee structure index",
        "Coffee density in 30 x 30m plot",
        "Coffee dominance",
    ]

    # Normalize radar values
    scaler = MinMaxScaler()
    df_scaled = df.copy()
    df_scaled[radar_cols] = scaler.fit_transform(df[radar_cols])

    site_options = df["Site ID"].tolist()
    return df, df_scaled, radar_cols, site_options


@app.cell
def _(mo, site_options):
    # --- Sidebar / UI controls
    mo.md("# Radar Chart – Site Comparison")
    return


@app.cell
def _(mo, site_options):
    selected_sites = mo.ui.multiselect(
        options=site_options,
        value=[site_options[0], site_options[1]],
        label="Select sites to compare",
    )
    selected_sites
    return (selected_sites,)


@app.cell
def _(cm, df, df_scaled, go, mcolors, mo, radar_cols, selected_sites):
    # --- Colour logic (relative shading by yield)
    _cmap = cm.get_cmap("Blues", 256)

    def get_color(rank, total):
        norm = rank / max(total - 1, 1)
        return mcolors.to_hex(_cmap(0.3 + 0.7 * norm))

    # --- Sort selected sites by yield
    selected_df = df[df["Site ID"].isin(selected_sites.value)].copy()
    selected_df = selected_df.sort_values("Mean_CC_Yield")
    n = len(selected_df)

    # --- Build radar figure
    fig = go.Figure()

    for i, (_, row) in enumerate(selected_df.iterrows()):
        site = row["Site ID"]
        yield_value = row["Mean_CC_Yield"]
        row_scaled = df_scaled[df_scaled["Site ID"] == site]
        values = row_scaled[radar_cols].values.flatten().tolist()
        values += values[:1]
        labels = radar_cols + [radar_cols[0]]
        color = get_color(i, n)

        fig.add_trace(
            go.Scatterpolar(
                r=values,
                theta=labels,
                fill="toself",
                name=f"{site} (Yield: {yield_value:.0f})",
                line=dict(color=color, width=3),
                opacity=0.85,
                hovertemplate=(
                    "<b>%{theta}</b><br>"
                    "Normalized value: %{r:.2f}<extra></extra>"
                ),
            )
        )

    fig.update_layout(
        polar=dict(radialaxis=dict(visible=True, range=[0, 1])),
        showlegend=True,
        height=700,
    )

    mo.ui.plotly(fig)