Interview-Analysis/03_Sentiment_Analysis.py

import marimo

__generated_with = "0.18.3"
app = marimo.App(width="medium")


@app.cell
def _():
    import marimo as mo
    import pandas as pd
    from pathlib import Path

    TAGUETTE_EXPORT_DIR = Path('./data/transcripts/taguette_results')
    WORKING_DIR = Path('./data/processing/03_sentiment_analysis')

    if not WORKING_DIR.exists():
        WORKING_DIR.mkdir(parents=True)
    if not TAGUETTE_EXPORT_DIR.exists():
        TAGUETTE_EXPORT_DIR.mkdir(parents=True)
    return WORKING_DIR, mo, pd


@app.cell(hide_code=True)
def _(mo):
    mo.md(r"""
    # Phase 1: Individual interview analysis
    - Create sentiment matrices for each interview (document)
    - Save the intermediate results to file in the `WORKING_DIR`
    """)
    return


@app.cell
def _(pd):
    import numpy as np

    def create_sentiment_matrix(df, document_name, column_prefix='VT - |CT - ', row_prefix='_V-|_C-'):
        """
        Create a sentiment matrix for a specific document.

        Parameters:
        - df: DataFrame with columns ['document', 'tag', '_context', 'sentiment']
        - document_name: Name of the document to filter by

        Returns:
        - DataFrame representing the sentiment matrix
        """
        # Filter for the specific document
        doc_df = df[df['document'] == document_name].copy()

        # Filter for rows where the tag matches the sentiment prefixes (VT-/CT-)
        sentiment_rows = doc_df[
            doc_df['tag'].str.contains(column_prefix, na=False)
        ].copy()

        if sentiment_rows.empty:
            print(f"No sentiment data found for document: {document_name}")
            return pd.DataFrame()

        # Filter for rows with valid Voice/Character context
        valid_rows = sentiment_rows[
            sentiment_rows['_context'].notna() & 
            (sentiment_rows['_context'].str.contains(row_prefix, na=False))
        ].copy()

        if valid_rows.empty:
            print(f"No Voice/Character context found for document: {document_name}")
            return pd.DataFrame()

        # Create aggregation: group by Voice/Character (_context) and Theme (tag)
        # Sum sentiment scores for each combination
        matrix_data = valid_rows.groupby(['_context', 'tag'])['sentiment'].sum().reset_index()

        # Pivot to create the matrix
        matrix = matrix_data.pivot(index='_context', columns='tag', values='sentiment')

        # Fill NaN with 0 (no sentiment data for that combination)
        matrix = matrix.fillna(0)

        # Convert to integers for cleaner display
        matrix = matrix.astype(int)

        return matrix

    return (create_sentiment_matrix,)


@app.cell(hide_code=True)
def _(mo):
    mo.md(r"""
    ## Step 1.1: Voice Sample vs. Theme Sentiment Matrix

    For each interview (document), create a matrix where:
    - Rows represent the different Voices (based on '_V-' tags)
    - Columns represent the different VoiceThemes(based on 'VT -' tags)
    - Each cell contains the aggregated sentiment score (sum) for that Voice/Theme combination
    """)
    return


@app.cell
def _(WORKING_DIR, all_tags_df, create_sentiment_matrix, mo):

    # Create matrices for each unique document
    documents = all_tags_df['document'].unique()
    matrices = {}

    for doc in documents:
        print(f"\n{'='*60}")
        print(f"Document: {doc}")
        print('='*60)
        matrix = create_sentiment_matrix(all_tags_df, doc, column_prefix='VT - ', row_prefix='_V-')
        if not matrix.empty:
            matrices[doc] = matrix
            print(matrix)
        else:
            print("No matrix data available")

        # Save to CSV
        timestamp = mo.utils.get_timestamp(short=True)
        filename = WORKING_DIR / f"{doc.replace(' ', '_')}_voice_theme_matrix_{timestamp}.csv"
        matrix.to_csv(filename)
        print(f"Matrix saved to: {filename}")

    # Store matrices in a variable for further analysis
    matrices
    return


@app.cell(hide_code=True)
def _(mo):
    mo.md(r"""
    ## Step 1.2: Character Sample vs. Theme Sentiment Matrix

    For each interview (document), create a matrix where:
    - Rows represent the different Characters (based on  '_C-' tags)
    - Columns represent the different CharacterThemes (based on 'CT -' tags)
    - Each cell contains the aggregated sentiment score (sum) for that Character/Theme combination
    """)
    return


@app.cell(hide_code=True)
def _(mo):
    mo.md(r"""
    ## Step 1.3: Chase Brand Sentiment

    TODO: not sure we have enough supporting data for this yet
    """)
    return


@app.cell(hide_code=True)
def _(mo):
    mo.md(r"""
    ## Step 1.x: Save Matrices to Files

    Save the matrices to CSV files in the WORKING_DIR for intermediate storage. Include a short timestamp in the filename so we can track runs.
    """)
    return


@app.cell
def _():
    # Save the matrices to CSV files in the WORKING_DIR for intermediate storage. Include a short timestamp in the filename so we can track runs.
    return


@app.cell(hide_code=True)
def _(mo):
    mo.md(r"""
    # Phase 2: Overall Results

    Aggregate results of all the interviews into master matrices.
    """)
    return


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