How to create dictionary when reading csv file?

Question

I've got the following brute force code:

import csv
d = {}
with open(csv_filename, 'r') as csv_file:
    reader = csv.reader(csv_file)
    for row in reader:
        key = row[0]
        d[key] = np.array(
            [[float(x) for x in row[1:4]], [float(x) for x in row[4:7]],
             [float(x) for x in row[7:]]], dtype=float)

Is there a way to rewrite it to avoid [1:4], [4:7] etc?

Answer 1

If I understood correctly you just need to reshape the data.

key = row[0]
d[key] = np.reshape(np.array(row).astype(float), (3, 3))