Datasets:

Modalities:
Image
Formats:
parquet
Languages:
English
DOI:
Libraries:
Datasets
Dask
License:
plantorgans / utils /csv_constructor.py
jpodivin's picture
Initial commit
ae6e9e1
#!/bin/env python3
import json
import glob
import os
import pandas as pd
import argparse
def main(label_path):
unwanted = [
'parent_prediction', 'parent_annotation',
'last_created_by', 'completed_by',
'created_username', 'created_ago',
'project', 'updated_by',
'file_upload', 'comment_authors', 'meta',
'unresolved_comment_count', 'last_comment_updated_at',
'project', 'updated_by',
'file_upload', 'comment_authors', 'created_at', 'updated_at', 'is_labeled',
'inner_id', 'total_annotations', 'cancelled_annotations', 'total_predictions', 'comment_count']
label_files = [p for p in glob.glob(os.path.join(label_path, "*"))]
label_csv = []
for l in label_files:
with open(l) as label:
label = json.load(label)
for k in unwanted:
label.pop(k, None)
label['task'].pop(k, None)
label_csv.append(label)
label_csv = pd.DataFrame(label_csv)
label_csv = label_csv.drop(columns=['draft_created_at', 'lead_time', 'last_action'], errors='ignore')
label_csv.to_csv('labels.csv')
if __name__ == "__main__":
parser = argparse.ArgumentParser("labelconvertor")
parser.add_argument("label_path", type=str)
arguments = parser.parse_args()
main(arguments.label_path)