app.py

import gradio as gr
import pandas as pd
import seaborn as sns
import matplotlib.pyplot as plt
from io import StringIO

def upload_csv(csv_file):
    data = pd.read_csv(csv_file.name)
    analysis_results = analyze_data(data)
    return analysis_results

def create_histogram(data, column, ax):
    sorted_data = data[column].value_counts().sort_index()
    sorted_data.plot(kind='bar', color='#007ACC', ax=ax)
    ax.set_title(f"{column} Distribution")
    ax.set_xlabel(column)
    ax.set_ylabel("Frequency")

def calculate_points(data):
    points = data.copy()
    points.iloc[:, ::2] -= 1
    points.iloc[:, 1::2] = 5 - points.iloc[:, 1::2]
    return points

def calculate_sus_score(points):
    total_points = points.sum(axis=1)
    user_scores = total_points * 2.5
    sus_score = user_scores.mean()
    return sus_score, user_scores

def analyze_data(data):
    # Remove timestamp column
    data = data.drop(columns=['Timestamp'])

    # Calculate points
    points = calculate_points(data)

    # Calculate SUS score and individual user scores
    sus_score, user_scores = calculate_sus_score(points)
    user_scores_table = pd.DataFrame(user_scores, columns=["User Score"])

    # Histograms for each question
    fig, axes = plt.subplots(len(data.columns), figsize=(6, 6 * len(data.columns)))
    for idx, column in enumerate(data.columns):
        create_histogram(data, column, axes[idx])
    plt.tight_layout()
    plt.savefig("histograms.png")

    # Correlation matrix
    corr_matrix = data.corr()

    # Clustermap
    sns.set(font_scale=0.8)
    shortened_columns = [f"Q{idx + 1}" for idx in range(len(data.columns))]
    cluster_map = sns.clustermap(corr_matrix, annot=True, cmap="coolwarm", figsize=(10, 10), xticklabels=shortened_columns, yticklabels=shortened_columns)
    plt.setp(cluster_map.ax_heatmap.get_xticklabels(), rotation=45)
    plt.savefig("cluster_map.png")

    return sus_score, "histograms.png", "cluster_map.png", data.describe().T, user_scores_table

iface = gr.Interface(
    upload_csv,
    [gr.inputs.File(label="Upload CSV file")],
    [
        gr.outputs.Textbox(label="SUS Score"),
        gr.outputs.Image(label="Histograms", type='numpy'),
        gr.outputs.Image(label="Cluster Map", type='numpy'),
        gr.outputs.Dataframe(label="Descriptive Statistics", type='pandas'),
        gr.outputs.Dataframe(label="Individual User Scores", type='pandas')
    ],
    title="Survey Analysis",
    description="Upload a survey CSV file to get the System Usability Scale (SUS) score, histograms, cluster map, and individual user scores.",
)

if __name__ == "__main__":
    iface.launch(debug=True)