app.py

import marimo

__generated_with = "0.11.6"
app = marimo.App(width="medium")


@app.cell(hide_code=True)
def _(mo):
    mo.md(r"""# Diabetes Dataset Analysis""")
    return


@app.cell(hide_code=True)
def _():
    import marimo as mo
    import polars as pl
    return mo, pl


@app.cell(hide_code=True)
def _(mo):
    mo.accordion(
        {"Notes": """
    ## Dataset Column Notes

    > Only highlighted columns of interest 

    * Diabetes_binary: [ 0 (No diabetes) | 1 (Pre/diabetes) ]
    * HighBP: [ 0 (No High BP) | 1 (High BP) ]
    * HighChol: [ 0 (No High BP) | 1 (High BP) ]
    * Stroke: [ 0 (Never) | 1 (Had) ]
    * HeartDiseaseorAttack: [ 0 (No) | 1 (Yes) ]
    * Smoker: [ 0 (<100 cigs lifetime) | 1 (>100 cigs lifetime) 
    * HvyAlcohol: [ 0 (<14 🍺/week for men, <7 🍺/week for women) | 1 (otherwise) ]
        """}
    )
    return


@app.cell(hide_code=True)
def _(pl):
    dataset_raw = pl.read_csv("dataset/diabetes_binary_health_indicators_BRFSS2015.csv")
    dataset_prior_conditions = dataset_raw.select(["Diabetes_binary", "HighBP", "HighChol", "Stroke", "HeartDiseaseorAttack"])
    dataset_prior_conditions.head()
    return dataset_prior_conditions, dataset_raw


@app.cell
def _(mo):
    mo.md("""## Testing Classifiers""")
    return


@app.cell
def _(dataset_prior_conditions, mo, pl):
    from sklearn.naive_bayes import BernoulliNB
    from sklearn.tree import DecisionTreeClassifier
    from sklearn.model_selection import train_test_split
    from sklearn.metrics import accuracy_score, classification_report, confusion_matrix

    X_priors_NB, y_priors_NB = dataset_prior_conditions.select(pl.exclude("Diabetes_binary")), dataset_prior_conditions.select("Diabetes_binary")
    X_train_priors, X_test_priors, y_train_priors, y_test_priors = train_test_split(
        X_priors_NB, y_priors_NB, random_state=33, test_size=0.25
    )

    bnb = BernoulliNB()
    dtc = DecisionTreeClassifier()
    y_pred_priors = bnb.fit(X_train_priors, y_train_priors).predict(X_test_priors)
    y_pred_dtc = dtc.fit(X_train_priors, y_train_priors).predict(X_test_priors)
    mo.accordion(
        {
            "Bernoulli NB Metrics": f"""    
        Accuracy             : {accuracy_score(y_test_priors, y_pred_priors)}

        Confusion Matrix:

    ```
        {confusion_matrix(y_test_priors, y_pred_priors)}
    ```

        Classification Report: 

    ```
        {classification_report(y_test_priors, y_pred_priors)}
    ```
        """,
        "Decision Tree Classifier": f"""    
        Accuracy             : {accuracy_score(y_test_priors, y_pred_dtc)}

        Confusion Matrix:

    ```
        {confusion_matrix(y_test_priors, y_pred_dtc)}
    ```

        Classification Report: 

    ```
        {classification_report(y_test_priors, y_pred_dtc)}
    ```
        """})
    return (
        BernoulliNB,
        DecisionTreeClassifier,
        X_priors_NB,
        X_test_priors,
        X_train_priors,
        accuracy_score,
        bnb,
        classification_report,
        confusion_matrix,
        dtc,
        train_test_split,
        y_pred_dtc,
        y_pred_priors,
        y_priors_NB,
        y_test_priors,
        y_train_priors,
    )


@app.cell
def _(mo):
    mo.md(r"""Looks like Bernoulli Naive Bayes' performs better on this dataset, as even though the Decision Tree Classifier has a bit better accuracy, the other metrics do give a better score on the BNB overall.""")
    return


@app.cell
def _(mo):
    mo.md(r"""# Diabetes Predictor using BNB""")
    return


@app.cell
def _(mo):
    priors_predict = mo.md(
        '''
        Do you suffer from?

        * {high_bp} - High Blood Pressure
        * {high_chol} - High Cholesterol
        * {stroke} - Stroke
        * {heart_disease_or_attack} - Heart Disease or Attack 
        '''
    ).batch(high_bp=mo.ui.checkbox(), high_chol=mo.ui.checkbox(), stroke=mo.ui.checkbox(), heart_disease_or_attack=mo.ui.checkbox())
    priors_predict
    return (priors_predict,)


@app.cell
def _(bnb, mo, priors_predict):
    diabetes_or_not = bool(bnb.predict([[i.value for i in priors_predict.values()],]))
    prediction = None
    if diabetes_or_not:
        prediction = mo.md("Diabetes").callout(kind="danger")
    else: 
        prediction = mo.md("No Diabetes").callout(kind="success")
    prediction
    return diabetes_or_not, prediction


if __name__ == "__main__":
    app.run()