app-sudaporn-research-normalised/app.py

import numpy as np
import pandas as pd
import plotly.graph_objects as go
import streamlit as st

def process_and_plot_csv(df):
    """Processes the uploaded CSV file and creates interactive Plotly plots."""

    # Remove unnamed columns
    df = df.loc[:, ~df.columns.str.contains('^Unnamed')]

    # Drop rows where book names are missing
    df = df.dropna(subset=[df.columns[0]])

    # Extract book names
    book_names = df.iloc[:, 0].astype(str)

    # Extract only numeric cumulative chapter values
    cumulative_books = df.iloc[:, 2:].apply(pd.to_numeric, errors='coerce')

    # Drop rows where all cumulative values are NaN
    cumulative_books = cumulative_books.dropna(how='all')

    # Convert to numpy array
    cumulative_books = cumulative_books.values

    # Remove books with only a single cumulative value
    valid_books = [book for book in cumulative_books if np.count_nonzero(~np.isnan(book)) > 1]
    valid_book_names = [book_names.iloc[i] for i in range(len(cumulative_books)) if np.count_nonzero(~np.isnan(cumulative_books[i])) > 1]

    # Debug info
    st.write(f"Total Books Before Filtering: {len(cumulative_books)}, After Filtering: {len(valid_books)}")

    # Normalize to 100 points for each book
    target_points = 100
    normalised_books = []
    normalised_book_names = []
    change_values = []

    for i, book in enumerate(valid_books):
        book = np.array(book)
        book = book[~np.isnan(book)]  # Remove NaNs

        original_points = np.arange(len(book))  # Original chapter indices
        target_indices = np.linspace(0, len(book) - 1, target_points)  # Normalize to 100 points
        normalised = np.interp(target_indices, original_points, book)  # Interpolation

        normalised_books.append(normalised[:target_points])  # Ensure exactly 100 points
        normalised_book_names.append(valid_book_names[i])
        change_values.append(normalised[-1] - normalised[0])  # Store change from start to end

    # Separate books into positive and negative groups based on start vs. end value
    positive_books = []
    negative_books = []
    positive_names = []
    negative_names = []

    # Creating bins dynamically for the pie chart
    min_change = min(change_values)
    max_change = max(change_values)
    num_bins = 16  # Adjust this number for more or fewer bins

    bin_edges = np.linspace(min_change, max_change, num_bins + 1)
    book_change_categories = {f"Change: {bin_edges[i]:.1f} to {bin_edges[i+1]:.1f}": [] for i in range(num_bins)}

    positive_vs_negative = {"Positive Books": [], "Negative Books": []}

    for i, book in enumerate(normalised_books):
        start_value = book[0]
        final_value = book[-1]
        change = final_value - start_value

        # Classify as positive or negative
        if final_value >= start_value:
            positive_books.append(book)
            positive_names.append(normalised_book_names[i])
            positive_vs_negative["Positive Books"].append(normalised_book_names[i])
        else:
            negative_books.append(book)
            negative_names.append(normalised_book_names[i])
            positive_vs_negative["Negative Books"].append(normalised_book_names[i])

        # Assign book to appropriate bin in the pie chart
        for j in range(num_bins):
            if bin_edges[j] <= change < bin_edges[j + 1]:
                category = f"Change: {bin_edges[j]:.1f} to {bin_edges[j + 1]:.1f}"
                book_change_categories[category].append(normalised_book_names[i])
                break

    # Convert categories to count data
    category_counts = {key: len(value) for key, value in book_change_categories.items()}
    positive_vs_negative_counts = {key: len(value) for key, value in positive_vs_negative.items()}

    # Function to plot books
    def plot_books(books, names, title):
        fig = go.Figure()

        for i, book in enumerate(books):
            fig.add_trace(go.Scatter(
                x=np.arange(1, target_points + 1),
                y=book,
                mode='lines',
                name=names[i],
                hoverinfo="x+y+name"
            ))

        fig.update_layout(
            title=title,
            xaxis_title="Normalised Chapter Index (0 to 100)",
            yaxis_title="Cumulative Change",
            legend_title="Books",
            hovermode="x unified"
        )

        return fig

    # Display the plots
    st.write("### 📈 Overall Normalised Cumulative Chapter Changes")
    st.plotly_chart(plot_books(normalised_books, normalised_book_names, "All Books"))

    st.write("### 📗 Positive End Books (Started At or Above End Level)")
    st.plotly_chart(plot_books(positive_books, positive_names, "Positive End Books"))

    st.write("### 📕 Negative End Books (Ended Lower Than Start)")
    st.plotly_chart(plot_books(negative_books, negative_names, "Negative End Books"))

    # Create a Pie Chart for Change in Points (Start to End)
    fig_pie_change = go.Figure(
        data=[go.Pie(
            labels=list(category_counts.keys()),
            values=list(category_counts.values()),
            hole=0.3
        )]
    )

    fig_pie_change.update_layout(
        title="📊 Book Change Distribution (Start to End Difference)",
        clickmode='event+select'  # Enable click events
    )

    st.write("### 📊 Book Change Categories (Start vs. End Points)")
    st.plotly_chart(fig_pie_change, use_container_width=True)

    # Create a Pie Chart for Positive vs. Negative Books
    fig_pie_pos_neg = go.Figure(
        data=[go.Pie(
            labels=list(positive_vs_negative_counts.keys()),
            values=list(positive_vs_negative_counts.values()),
            hole=0.3,
            marker=dict(colors=["green", "red"])  # Green for positive, Red for negative
        )]
    )

    fig_pie_pos_neg.update_layout(
        title="📊 Positive vs. Negative Books",
        clickmode='event+select'  # Enable click events
    )

    st.write("### 📊 Positive vs. Negative Books")
    st.plotly_chart(fig_pie_pos_neg, use_container_width=True)

    # Dropdown for selecting book categories
    selected_category = st.selectbox("Select a category to see book names:", 
                                     ["None"] + list(book_change_categories.keys()) + list(positive_vs_negative.keys()))

    if selected_category != "None":
        st.write(f"### 📚 Books in {selected_category}:")
        if selected_category in book_change_categories:
            st.write(book_change_categories[selected_category])
        else:
            st.write(positive_vs_negative[selected_category])

# Streamlit UI
st.title("📖 Book Analysis: Cumulative Chapter Changes (Interactive)")

# File uploader
uploaded_file = st.file_uploader("📂 Upload a CSV File", type=["csv"])

if uploaded_file is not None:
    df = pd.read_csv(uploaded_file)  # Read uploaded file
    st.write("### 📊 Data Preview:")
    st.dataframe(df)  # Show uploaded data

    st.write("### 📈 Interactive Plots for Normalised Chapter Changes:")
    process_and_plot_csv(df)  # Process and plot the data