Currently I have the csv data in below format:
However I want the data to be in spacy (version 2.0) format, where it should be grouped by "data" column in below format:
I have tried it to group by "data" field using code below:
def data_format():
df = pd.read_csv("Training_data_sample.csv")
grouped = df.groupby(['data'])
df_1 = pd.DataFrame(grouped)
df_1.to_csv("Test_format.csv", encoding="utf-8",index=False)
try:
df
data positions field
0 data1 (37,41) fielf_1
1 data1 (4,14) fielf_2
2 data1 (88,121) fielf_3
3 data1 (227,267) fielf_4
4 data2 (37,41) fielf_1
5 data2 (4,14) fielf_2
6 data2 (88,121) fielf_3
7 data2 (227,267) fielf_4
df['n'] = df.apply(lambda x: (x['positions'], x['field']), axis=1)
df.groupby('data')['n'].apply(list).reset_index()
data n
0 data1 [((37,41), fielf_1), ((4,14), fielf_2), ((88,1...
1 data2 [((37,41), fielf_1), ((4,14), fielf_2), ((88,1...
#if you need the result in dictionary format:
df.groupby('data')['n'].apply(list).to_dict()
{'data1': [('(37,41)', 'fielf_1'), ('(4,14)', 'fielf_2'), ('(88,121)', 'fielf_3'), ('(227,267)', 'fielf_4')],
'data2': [('(37,41)', 'fielf_1'), ('(4,14)', 'fielf_2'), ('(88,121)', 'fielf_3'), ('(227,267)', 'fielf_4')]}