Convert JSON output from V7 Darwin to csv.

extract_darwin_to_csv.py

"""
Convert JSON output from V7 Darwin to csv.
"""
import json
import os.path
import pandas as pd

filenames = [f"{name}.json" for name in list('abcde')]
data_dir = "data/lizzie-reader-study-test-part-1"


rows = []
for filename in filenames:
    path = os.path.join(data_dir, filename)
    with open(path) as f:
        data = json.load(f)
    for annotation in data["annotations"]:
        for attribute in annotation["attributes"]:
            rows.append((
                data["dataset"],
                data["image"]["filename"],
                annotation["annotators"][0]["full_name"],
                annotation["name"],
                attribute
            ))

df = pd.DataFrame(data=rows, columns=["dataset", "filename", "annotator", "annotation_name", "attribute"])

df.to_csv("lizzie-reader-study-test-part-1.csv", index=False)