Why they use collections in writing csv?) Puzzles me

Question

Came across this topic Save results to csv file with Python

All i needed - to write csv changes to file. BUT. This code stole some of my rows)) unreasonable amount(instead of five as in code)

Could you please explain why do they use collections in this simple action? And why is counter used here?

Use csv.writer:

import csv

with open('thefile.csv', 'rb') as f:
  data = list(csv.reader(f))

import collections
counter = collections.defaultdict(int)
for row in data:
    counter[row[0]] += 1


writer = csv.writer(open("/path/to/my/csv/file", 'w'))
for row in data:
    if counter[row[0]] >= 4:
        writer.writerow(row)

Answer 1

Not sure what is going on in the original code, but I think it reads the input and creates a dictionary of the data lines and assigns a counter to each line. So this wouldn't work correctly if there are two identical data lines btw. And then it uses this dictionary to determine which are the first 4 lines, in order to skip them when writing to the output (which is not a good use of dictionaries imho).

if counter[row[0]] >= 4:

The main thing is that it's an old answer from 11 years ago. Back then Python libraries like pandas for handling csv files weren't as commonly used. Now it's easier to just do something like this.

import pandas as pd

# semicolon separated and comma decimals, skip first 4 rows
df_in = pd.read_csv("oldfile.csv", delimiter=";", decimal=",", skiprows=(0,1,2,3))

# comma separated and point decimals
df_in.to_csv("newfile.csv", sep=",", decimal=".")

Answer 2

This row

counter = collections.defaultdict(int)

creates a defaultdict where the values are integers, and the default value is zero. Then this bit

for row in data:
    counter[row[0]] += 1

scans the input, and counts how many times the value in the first field occurs in the file. Finally, the code outputs only those rows whose first value occurs at least 4 times. This could all be made a bit shorter by using collections.Counter instead, but I don't know whether that was in the standard library ten years ago.