app.py

import os  # For filesystem operations
import shutil  # For directory cleanup
import zipfile  # For extracting model archives
import pathlib  # For path manipulations
import pandas  # For tabular data handling
import gradio  # For interactive UI
import huggingface_hub  # For downloading model assets
import autogluon.tabular  # For loading and running AutoGluon predictors

# Settings
MODEL_REPO_ID = "its-zion-18/flowers-tabular-autolguon-predictor"
ZIP_FILENAME  = "autogluon_predictor_dir.zip"
CACHE_DIR = pathlib.Path("hf_assets")
EXTRACT_DIR = CACHE_DIR / "predictor_native"

# Feature column names and target column names
FEATURE_COLS = [
    "flower_diameter_cm",
    "petal_length_cm",
    "petal_width_cm",
    "petal_count",
    "stem_height_cm"
]
TARGET_COL = "color"


# Encoding for outcome questions

OUTCOME_LABELS = {
    0: "Pink",
    1: "Red",
    2: "Orange",
    3: "White",
    4: "Yellow",
    5: "Purple",
}

# Download & load the native predictor
def _prepare_predictor_dir() -> str:
    CACHE_DIR.mkdir(parents=True, exist_ok=True)
    local_zip = huggingface_hub.hf_hub_download(
        repo_id=MODEL_REPO_ID,
        filename=ZIP_FILENAME,
        repo_type="model",
        local_dir=str(CACHE_DIR),
        local_dir_use_symlinks=False,
    )
    if EXTRACT_DIR.exists():
        shutil.rmtree(EXTRACT_DIR)
    EXTRACT_DIR.mkdir(parents=True, exist_ok=True)
    with zipfile.ZipFile(local_zip, "r") as zf:
        zf.extractall(str(EXTRACT_DIR))
    contents = list(EXTRACT_DIR.iterdir())
    predictor_root = contents[0] if (len(contents) == 1 and contents[0].is_dir()) else EXTRACT_DIR
    return str(predictor_root)

PREDICTOR_DIR = _prepare_predictor_dir()
PREDICTOR = autogluon.tabular.TabularPredictor.load(PREDICTOR_DIR, require_py_version_match=False)

# A mapping utility to make it easier to encode the variables
def _human_label(c):
    try:
        ci = int(c)
        if ci in OUTCOME_LABELS:
            return OUTCOME_LABELS[ci]
    except Exception:
        pass
    if c in OUTCOME_LABELS:
        return OUTCOME_LABELS[c]
    return str(c)

# This functions takes all of our features,e ncodes this accordingly, and performs a predictions
def do_predict(diameter, length, width, number, height):

    row = {
        FEATURE_COLS[0]: float(diameter),
        FEATURE_COLS[1]: float(length),
        FEATURE_COLS[2]: float(width),
        FEATURE_COLS[3]: int(number),
        FEATURE_COLS[4]: float(height),
    }
    X = pandas.DataFrame([row], columns=FEATURE_COLS)

    pred_series = PREDICTOR.predict(X)
    raw_pred = pred_series.iloc[0]

    try:
        proba = PREDICTOR.predict_proba(X)
        if isinstance(proba, pandas.Series):
            proba = proba.to_frame().T
    except Exception:
        proba = None

    pred_label = _human_label(raw_pred)

    proba_dict = None
    if proba is not None:
        row0 = proba.iloc[0]
        tmp = {}
        for cls, val in row0.items():
            key = _human_label(cls)
            tmp[key] = float(val) + float(tmp.get(key, 0.0))
        proba_dict = dict(sorted(tmp.items(), key=lambda kv: kv[1], reverse=True))

    df_out = pandas.DataFrame([{
        "Predicted outcome": pred_label,
        "Confidence (%)": round((proba_dict.get(pred_label, 1.0) if proba_dict else 1.0) * 100, 2),
    }])

    md = f"**Prediction:** {pred_label}"
    if proba_dict:
        md += f"  \n**Confidence:** {round(proba_dict.get(pred_label, 0.0) * 100, 2)}%"

    return proba_dict

# Representative examples
EXAMPLES = [
    [3.4, 1.3, 1,  7,      68.7],
    [5.7, 3.2, 0.9, 5,     21.2],
    [6.9, 3.8, 0.6, 17, 64.2],
    [5.6, 1, 1.5,   7,      73.4, ],
    [7.7, 1.5, 1.8, 13, 73.8],
]

# Gradio UI
with gradio.Blocks() as demo:
    # Provide an introduction
    gradio.Markdown("# What Color is the Flower?")
    gradio.Markdown("""
    This is a simple app that uses the model at its-zion-18/flowers-tabular-autolguon-predictor
    to predict the color of a flower based off its characteristics, utilizing data found at
    scottymcgee/flowers. To use the interface, make 
    selections using the interface elements shown below. 
    """)

    with gradio.Row():
        diameter = gradio.Slider(3, 10, step=0.1, value=5.0, label=FEATURE_COLS[0])
        length = gradio.Slider(1, 5, step=0.1, value=3.0, label=FEATURE_COLS[1])
        width = gradio.Slider(0.5, 2, step=0.1, value=1.0, label=FEATURE_COLS[2])

    with gradio.Row():
        number = gradio.Slider(5, 30, step=1, value=15, label=FEATURE_COLS[3])
        height = gradio.Slider(10, 80, step=0.1, value=50.0, label=FEATURE_COLS[4])

    proba_pretty = gradio.Label(num_top_classes=5, label="Class probabilities")

    inputs = [diameter, length, width, number, height]
    for comp in inputs:
        comp.change(fn=do_predict, inputs=inputs, outputs=[proba_pretty])

    gradio.Examples(
        examples=EXAMPLES,
        inputs=inputs,
        label="Representative examples",
        examples_per_page=5,
        cache_examples=False,
    )

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