Reads a CSV of topic words and weights stored in one row across 11 columns. Creates a new CSV with the topics words and scores stored across four rows and seven columns.

input.csv

lda-data-formatter.py

import csv

# Open the input CSV file
with open('input.csv', 'r') as input_file:
    csv_reader = csv.reader(input_file)

    # Create a dictionary to store the data pairs by ID
    id_data_pairs = {}

    # Loop through each row in the input CSV file
    for row in csv_reader:
        # Extract the ID value from the first column of the input row
        id_value = row[0]

        # Split the remaining columns into pairs of data
        data_pairs = [row[i:i+2] for i in range(1, len(row), 2)]

        # Add the data pairs to the dictionary, grouped by ID
        if id_value in id_data_pairs:
            id_data_pairs[id_value] += data_pairs
        else:
            id_data_pairs[id_value] = data_pairs

    # Create a list to store the output rows
    output_rows = []

    # Loop through each ID in the dictionary
    for id_value in id_data_pairs.keys():
        # Get the data pairs for the current ID
        data_pairs = id_data_pairs[id_value]

        # Determine the number of pairs and the number of rows needed
        num_pairs = len(data_pairs)
        num_rows = num_pairs // 3
        if num_pairs % 3 != 0:
            num_rows += 1

        # Create a list of empty rows
        rows = [[] for i in range(num_rows)]

        # Loop through the data pairs and add them to the rows list
        for i, pair in enumerate(data_pairs):
            row_index = i % num_rows
            rows[row_index].extend(pair)

        # Add the rows to the output list, with the ID value in the first column of each row
        for row in rows:
            output_rows.append([id_value] + row)

    # Open a new CSV file for writing the output
    with open('output.csv', 'w', newline='') as output_file:
        csv_writer = csv.writer(output_file)

        # Write the output rows to the output CSV file
        for row in output_rows:
            csv_writer.writerow(row)

output.csv