File size: 1,881 Bytes
9965bf6 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 |
import glob
import pandas as pd
from collections import defaultdict
from argparse import ArgumentParser
parser = ArgumentParser()
parser.add_argument('--path_to_folder', required=True, type=str, default='loggings/eval_uncond/ours/',
help='Path (absolute) to the first dataset (folder)')
args = parser.parse_args()
# whether to use OA or KL
metric = 'OA'
# attribute to look at
attributes = ['total_used_pitch',
'avg_IOI',
'total_pitch_class_histogram',
'pitch_range',
'mean_note_velocity',
'mean_note_duration',
'note_density',
'avg'
]
csv_files = glob.glob(args.path_to_folder + '*.csv')
results_mean = defaultdict(list)
results_std = defaultdict(list)
# This will print out all paths that end with '.csv'
for file_path in csv_files:
file_name = file_path.split('/')[-1]
dataset = file_name.split('.')[0]
method = file_name.split('.')[1]
stat_type = file_name.split('.')[2]
df = pd.read_csv(file_path)
if stat_type == 'mean':
results_mean["dataset"].append(dataset)
results_mean["method"].append(method)
for attr in attributes:
results_mean[attr].append(df[df.attribute == attr][metric].item())
else:
results_std["dataset"].append(dataset)
results_std["method"].append(method)
for attr in attributes:
results_std[attr].append(df[df.attribute == attr][metric].item())
df_results_mean = pd.DataFrame(results_mean)
sorted_df = df_results_mean.sort_values(by=['dataset', 'method'])
sorted_df.to_csv(args.path_to_folder + 'results_mean.csv', index=False)
df_results_std = pd.DataFrame(results_std)
sorted_df = df_results_std.sort_values(by=['dataset', 'method'])
sorted_df.to_csv(args.path_to_folder + 'results_std.csv', index=False)
print("done")
|