create_dendrogram.py

import json
import csv


def get_min_distance(m):
    # Initiate variables that will keep track of the minimum value in the distance matrix and its indices
    min_dist = float('inf')
    min_i, min_j = None, None

    # Loop through distance matrix and update minimum value and indices
    for i in range(len(m)):
        for j in range(len(m[i])):
            if m[i][j] < min_dist:
                min_dist = m[i][j]
                min_i, min_j = i, j

    # Return indices of the minimum value in the distance matrix
    return min_i, min_j


def regenerate_matrix(m, d):
    # Find minimum value in distance matrix
    x, y = get_min_distance(m)
    dist = m[x][y] / 2

    # Identify indices and data that are not affected by new clustering
    old_idx = [idx for idx in range(len(d)) if idx not in (x, y)]
    children = [entry for idx, entry in enumerate(d) if idx in (x, y)]

    # Calculate new row in distance matrix after clustering
    new_name = d[x]['name'] + '|' + d[y]['name']
    new_row = []

    for idx in old_idx:
        x_dist = m[x][idx] if idx < x else m[idx][x]
        y_dist = m[y][idx] if idx < y else m[idx][y]

        avg_dist = (x_dist + y_dist) / 2
        new_row.append(avg_dist)

    # Keep rows in distance matrix that are not affected by new clustering
    d = [entry for idx, entry in enumerate(d) if idx in old_idx]
    m = [[d for i, d in enumerate(row) if i in old_idx] for idx, row in enumerate(m) if idx in old_idx]

    # Add new row to the matrix and data
    d.append({'name': new_name, 'distance': dist, 'children': children})
    m.append(new_row)

    return m, d


def get_json(file, tab=False):
    # Input is a tab-delimited or comma-separated distance matrix
    if tab:
        reader = csv.reader(file.splitlines(), delimiter='\t')
    else:
        reader = csv.reader(file)

    matrix = []  # Stores distance matrix after each round of clustering
    data = []  # Eventually contains final clustering using UPGMA method

    # Give matrix and data initial values from input
    for idx, row in enumerate(reader):
        data.append({'name': row.pop(0), 'distance': 0})
        matrix.append([float(i) for i in row[:idx]])

    # Update matrix and data after each round of clustering
    while len(data) > 1:
        matrix, data = regenerate_matrix(matrix, data)

    # Return final clustering contained in data to be visualized
    return json.dumps(data[0])