зеркало из
https://github.com/iharh/notes.git
synced 2025-11-05 08:06:11 +02:00
38 строки
954 B
Plaintext
38 строки
954 B
Plaintext
http://pandas.pydata.org/pandas-docs/stable/groupby.html
|
|
https://www.shanelynn.ie/summarising-aggregation-and-grouping-data-in-python-pandas/
|
|
https://www.datacamp.com/community/tutorials/pandas-split-apply-combine-groupby
|
|
https://www.datacamp.com/community/tutorials/pandas-multi-index
|
|
|
|
groupby(...
|
|
as_index=False,
|
|
sort=True,
|
|
level=0,
|
|
level='second',
|
|
level=['first', 'second'],
|
|
).reset_index()
|
|
|
|
for name, group in dfallgrouped:
|
|
print(name)
|
|
print(group.iloc[2:-2])
|
|
|
|
dfallgrouped.apply(lambda grp: grp.iloc[2:-2])
|
|
|
|
.groups
|
|
.get_group(...)
|
|
.first()
|
|
.mean()
|
|
.aggregate(np.sum)
|
|
.sum()
|
|
.sum().reset_index()
|
|
.agg([np.sum, np.mean, np.std])
|
|
# agg by one column
|
|
groupby(...)['C'].agg({'sum':np.sum, 'mean':np.mean})
|
|
# ... by distinct columns
|
|
groupby(...).agg({'sum':np.sum, 'mean':np.mean})
|
|
|
|
mean = lambda x: (x-np.mean(x))
|
|
.transform(mean)
|
|
|
|
df.loc[df.groupby('A')['B'].idxmin()]
|
|
df.sort('B').groupby('A'), as_index=False).first()
|