app.py

from pathlib import Path

import gradio as gr
import pandas as pd
import plotly.express as px
import plotly.graph_objects as go


CSV_PATH = Path("metrics.csv")


def load_data() -> pd.DataFrame:
    df = pd.read_csv(CSV_PATH)
    # Normalize/parse columns
    if "date" not in df.columns:
        raise ValueError("Expected a 'date' column in metrics.csv")
    df["date"] = pd.to_datetime(df["date"], errors="coerce")
    df = df.sort_values("date").reset_index(drop=True)
    # Ensure numeric columns are numeric
    for c in df.columns:
        if c == "date":
            continue
        df[c] = pd.to_numeric(df[c], errors="coerce")
    return df


def line_figure(df: pd.DataFrame, cols: list[str], title: str, yaxis_title: str = "") -> go.Figure:
    if not cols:
        # Empty placeholder so the UI doesn't error
        fig = go.Figure()
        fig.update_layout(title=f"{title} (no series selected)")
        return fig
    fig = px.line(
        df,
        x="date",
        y=cols,
        markers=True,
        title=title,
    )
    # Improve layout for time series
    fig.update_layout(
        legend_title_text="Series",
        xaxis_title="Date",
        yaxis_title=yaxis_title,
        hovermode="x unified",
        margin={"l": 50, "r": 20, "t": 50, "b": 40},
    )
    return fig


TAB_SPEC = {
    "Docstrings": [
        "docstring coverage",
        "docstring missing",
    ],
    "Size (Lines/Statements/Expressions/Parameters)": [
        "lines mean",
        "lines max",
        "lines 90th-percentile",
        "statements mean",
        "statements max",
        "statements 90th-percentile",
        "expressions mean",
        "expressions max",
        "expressions 90th-percentile",
        "parameters mean",
        "parameters max",
        "parameters 90th-percentile",
    ],
    "Complexity": [
        "cyclomatic_complexity mean",
        "cyclomatic_complexity max",
        "cyclomatic_complexity 90th-percentile",
    ],
    "Typing": [
        "type_coverage mean",
        "type_coverage min",
        "type_coverage 50th-percentile",
    ],
    "Duplication": [
        "duplication.score mean",
        "duplication.score max",
        "duplication.score 90th-percentile",
        "duplication.score 50th-percentile",
        "duplication.duplicated-lines total",
    ],
    "TODOs": [
        "todo_comments total",
    ],
    "CLOC (Repository scope)": [
        "files",
        "lines blank",
        "lines comment",
        "lines code",
    ],
}

Y_LABELS = {
    "Docstrings": "value",
    "Size (Lines/Statements/Expressions/Parameters)": "count",
    "Complexity": "complexity",
    "Typing": "fraction / coverage",
    "Duplication": "score / lines",
    "TODOs": "count",
    "CLOC (Repository scope)": "lines / files",
}

DF = load_data()

with gr.Blocks(title="Code Metrics – Time Series", fill_height=True) as demo:
    gr.Markdown(
        "## Diffusers Code Metrics Over Time\n"
        f"Loaded **{CSV_PATH}** with {len(DF)} rows spanning "
        f"{DF['date'].min().date()} → {DF['date'].max().date()}.\n\n"
        "Use each tab to pick the series you want to plot."
    )

    with gr.Tabs():
        tab_controls = []
        tab_plots = []

        for tab_name, series in TAB_SPEC.items():
            available = [s for s in series if s in DF.columns]  # guard against missing cols
            with gr.Tab(tab_name):
                with gr.Row():
                    sel = gr.CheckboxGroup(
                        choices=available,
                        value=available,
                        label="Series",
                    )
                plot = gr.Plot(
                    value=line_figure(DF, available, tab_name, Y_LABELS.get(tab_name, "")),
                    show_label=False,
                )

                sel.change(
                    fn=lambda cols, t=tab_name: line_figure(DF, cols, t, Y_LABELS.get(t, "")),
                    inputs=sel,
                    outputs=plot,
                )

                tab_controls.append(sel)
                tab_plots.append(plot)


if __name__ == "__main__":
    demo.launch()