File size: 615 Bytes
0c1f7b6
 
 
 
 
 
 
dcf770c
 
 
 
 
 
 
 
0c1f7b6
5c5c1ed
 
0c1f7b6
5c5c1ed
0c1f7b6
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
import pandas as pd
from datasets import load_dataset

data = load_dataset('relbert/semeval2012_relational_similarity_v3')
stats = []
for k in data.keys():
    for i in data[k]:
        stats.append(
            {
                'relation_type': i['relation_type'],
                'split': k,
                'positives': len(i['positives']),
                'negatives': len(i['negatives']),
                'level': i['level']
            })
df = pd.DataFrame(stats)
g = df.groupby(['relation_type', 'level', 'split']).sum()
g.to_csv('stats.csv')
with open('stats.md', 'w') as f:
    f.write(g.to_markdown())