Pandas to_csv progress bar with tqdm

user_11918269 0
Viewed 2062

As the title suggests, I am trying to display a progress bar while performing pandas.to_csv.
I have the following script:

def filter_pileup(pileup, output, lists):
    tqdm.pandas(desc='Reading, filtering, exporting', bar_format=BAR_DEFAULT_VIEW)
    # Reading files
    pileup_df = pd.read_csv(pileup, '\t', header=None).progress_apply(lambda x: x)
    lists_df = pd.read_csv(lists, '\t', header=None).progress_apply(lambda x: x)
    # Filtering pileup
    intersection = pd.merge(pileup_df, lists_df, on=[0, 1]).progress_apply(lambda x: x)
    intersection.columns = [i for i in range(len(intersection.columns))]
    intersection = intersection.loc[:, 0:5]
    # Exporting filtered pileup
    intersection.to_csv(output, header=None, index=None, sep='\t')

On the first few lines I have found a way to integrate a progress bar but this method doesn't work for the last line, How can I achieve that?

tqdm pandas python
1 Answers

You can divide the dataframe into chunks of n rows and save the dataframe to a csv chunk by chunk using mode='w' for the first row and mode="a" for the rest:

Example:

import numpy as np
import pandas as pd
from tqdm import tqdm

df = pd.DataFrame(data=[i for i in range(0, 10000000)], columns = ["integer"])

print(df.head(10))

chunks = np.array_split(df.index, 100) # split into 100 chunks

for chunck, subset in enumerate(tqdm(chunks)):
    if chunck == 0: # first row
        df.loc[subset].to_csv('data.csv', mode='w', index=True)
    else:
        df.loc[subset].to_csv('data.csv', header=None, mode='a', index=True)

Output:

    integer
0        0
1        1
2        2
3        3
4        4
5        5
6        6
7        7
8        8
9        9

100%|██████████| 100/100 [00:12<00:00,  8.12it/s]
user_13552870
Related