Skip to content

Commit 985ebd4

Browse files
pandrespandres
committed
Add scaffold docstring for groupby.DataFrameGroupBy.resample.
1 parent aedbd94 commit 985ebd4

File tree

1 file changed

+162
-2
lines changed

1 file changed

+162
-2
lines changed

‎pandas/core/groupby.py

Lines changed: 162 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -1475,9 +1475,169 @@ def describe(self, **kwargs):
14751475
@Substitution(name='groupby')
14761476
@Appender(_doc_template)
14771477
def resample(self, rule, *args, **kwargs):
1478-
"""
1479-
Provide resampling when using a TimeGrouper
1478+
"""Provide resampling when using a TimeGrouper.
1479+
14801480
Return a new grouper with our resampler appended
1481+
1482+
Parameters
1483+
----------
1484+
rule : str
1485+
Medio.
1486+
args
1487+
Hola.
1488+
kwargs
1489+
Chau.
1490+
1491+
Returns
1492+
-------
1493+
Grouper
1494+
Return a new grouper with our resampler appended
1495+
1496+
Examples
1497+
--------
1498+
1499+
Start by creating a series with 9 one minute timestamps.
1500+
>>> index = pd.date_range('1/1/2000', periods=9, freq='T')
1501+
>>> series = pd.Series(range(9), index=index)
1502+
>>> series
1503+
2000-01-01 00:00:00 0
1504+
2000-01-01 00:01:00 1
1505+
2000-01-01 00:02:00 2
1506+
2000-01-01 00:03:00 3
1507+
2000-01-01 00:04:00 4
1508+
2000-01-01 00:05:00 5
1509+
2000-01-01 00:06:00 6
1510+
2000-01-01 00:07:00 7
1511+
2000-01-01 00:08:00 8
1512+
Freq: T, dtype: int64
1513+
1514+
Downsample the series into 3 minute bins and sum the values
1515+
of the timestamps falling into a bin.
1516+
>>> series.resample('3T').sum()
1517+
2000-01-01 00:00:00 3
1518+
2000-01-01 00:03:00 12
1519+
2000-01-01 00:06:00 21
1520+
Freq: 3T, dtype: int64
1521+
1522+
Downsample the series into 3 minute bins as above, but label each
1523+
bin using the right edge instead of the left. Please note that the
1524+
value in the bucket used as the label is not included in the bucket,
1525+
which it labels. For example, in the original series the
1526+
bucket ``2000-01-01 00:03:00`` contains the value 3, but the summed
1527+
value in the resampled bucket with the label ``2000-01-01 00:03:00``
1528+
does not include 3 (if it did, the summed value would be 6, not 3).
1529+
To include this value close the right side of the bin interval as
1530+
illustrated in the example below this one.
1531+
>>> series.resample('3T', label='right').sum()
1532+
2000-01-01 00:03:00 3
1533+
2000-01-01 00:06:00 12
1534+
2000-01-01 00:09:00 21
1535+
Freq: 3T, dtype: int64
1536+
1537+
Downsample the series into 3 minute bins as above, but close the right
1538+
side of the bin interval.
1539+
>>> series.resample('3T', label='right', closed='right').sum()
1540+
2000-01-01 00:00:00 0
1541+
2000-01-01 00:03:00 6
1542+
2000-01-01 00:06:00 15
1543+
2000-01-01 00:09:00 15
1544+
Freq: 3T, dtype: int64
1545+
1546+
Upsample the series into 30 second bins.
1547+
>>> series.resample('30S').asfreq()[0:5] #select first 5 rows
1548+
2000-01-01 00:00:00 0.0
1549+
2000-01-01 00:00:30 NaN
1550+
2000-01-01 00:01:00 1.0
1551+
2000-01-01 00:01:30 NaN
1552+
2000-01-01 00:02:00 2.0
1553+
Freq: 30S, dtype: float64
1554+
1555+
Upsample the series into 30 second bins and fill the ``NaN``
1556+
values using the ``pad`` method.
1557+
>>> series.resample('30S').pad()[0:5]
1558+
2000-01-01 00:00:00 0
1559+
2000-01-01 00:00:30 0
1560+
2000-01-01 00:01:00 1
1561+
2000-01-01 00:01:30 1
1562+
2000-01-01 00:02:00 2
1563+
Freq: 30S, dtype: int64
1564+
1565+
Upsample the series into 30 second bins and fill the
1566+
``NaN`` values using the ``bfill`` method.
1567+
>>> series.resample('30S').bfill()[0:5]
1568+
2000-01-01 00:00:00 0
1569+
2000-01-01 00:00:30 1
1570+
2000-01-01 00:01:00 1
1571+
2000-01-01 00:01:30 2
1572+
2000-01-01 00:02:00 2
1573+
Freq: 30S, dtype: int64
1574+
1575+
Pass a custom function via ``apply``
1576+
>>> def custom_resampler(array_like):
1577+
... return np.sum(array_like)+5
1578+
>>> series.resample('3T').apply(custom_resampler)
1579+
2000-01-01 00:00:00 8
1580+
2000-01-01 00:03:00 17
1581+
2000-01-01 00:06:00 26
1582+
Freq: 3T, dtype: int64
1583+
1584+
For a Series with a PeriodIndex, the keyword `convention` can be
1585+
used to control whether to use the start or end of `rule`.
1586+
>>> s = pd.Series([1, 2], index=pd.period_range('2012-01-01', freq='A', periods=2))
1587+
>>> s
1588+
2012 1
1589+
2013 2
1590+
Freq: A-DEC, dtype: int64
1591+
1592+
Resample by month using 'start' `convention`. Values are assigned to
1593+
the first month of the period.
1594+
>>> s.resample('M', convention='start').asfreq().head()
1595+
2012-01 1.0
1596+
2012-02 NaN
1597+
2012-03 NaN
1598+
2012-04 NaN
1599+
2012-05 NaN
1600+
Freq: M, dtype: float64
1601+
1602+
Resample by month using 'end' `convention`. Values are assigned to
1603+
the last month of the period.
1604+
>>> s.resample('M', convention='end').asfreq()
1605+
2012-12 1.0
1606+
2013-01 NaN
1607+
2013-02 NaN
1608+
2013-03 NaN
1609+
2013-04 NaN
1610+
2013-05 NaN
1611+
2013-06 NaN
1612+
2013-07 NaN
1613+
2013-08 NaN
1614+
2013-09 NaN
1615+
2013-10 NaN
1616+
2013-11 NaN
1617+
2013-12 2.0
1618+
Freq: M, dtype: float64
1619+
1620+
For DataFrame objects, the keyword ``on`` can be used to specify the
1621+
column instead of the index for resampling.
1622+
>>> df = pd.DataFrame(data=9*[range(4)], columns=['a', 'b', 'c', 'd'])
1623+
>>> df['time'] = pd.date_range('1/1/2000', periods=9, freq='T')
1624+
>>> df.resample('3T', on='time').sum()
1625+
a b c d
1626+
time
1627+
2000-01-01 00:00:00 0 3 6 9
1628+
2000-01-01 00:03:00 0 3 6 9
1629+
2000-01-01 00:06:00 0 3 6 9
1630+
1631+
For a DataFrame with MultiIndex, the keyword ``level`` can be used to
1632+
specify on level the resampling needs to take place.
1633+
>>> time = pd.date_range('1/1/2000', periods=5, freq='T')
1634+
1635+
>>> df2 = pd.DataFrame(data=10*[range(4)], columns=['a', 'b', 'c', 'd'], index=pd.MultiIndex.from_product([time, [1, 2]]) )
1636+
>>> df2.resample('3T', level=0).sum()
1637+
a b c d
1638+
2000-01-01 00:00:00 0 6 12 18
1639+
2000-01-01 00:03:00 0 4 8 12
1640+
14811641
"""
14821642
from pandas.core.resample import get_resampler_for_grouping
14831643
return get_resampler_for_grouping(self, rule, *args, **kwargs)

0 commit comments

Comments
 (0)