By “group by” we are referring to a process involving one or more of the following steps:
Splitting the data into groups based on some criteria.
Applying a function to each group independently.
Combining the results into a data structure.
Out of these, the split step is the most straightforward. In fact, in many situations we may wish to split the data set into groups and do something with those groups. In the apply step, we might wish to do one of the following:
Aggregation: compute a summary statistic (or statistics) for each group. Some examples:
Compute group sums or means. Compute group sizes / counts.
Compute group sums or means.
Compute group sizes / counts.
Transformation: perform some group-specific computations and return a like-indexed object. Some examples:
Standardize data (zscore) within a group. Filling NAs within groups with a value derived from each group.
Standardize data (zscore) within a group.
Filling NAs within groups with a value derived from each group.
Filtration: discard some groups, according to a group-wise computation that evaluates True or False. Some examples:
Discard data that belongs to groups with only a few members. Filter out data based on the group sum or mean.
Discard data that belongs to groups with only a few members.
Filter out data based on the group sum or mean.
Some combination of the above: GroupBy will examine the results of the apply step and try to return a sensibly combined result if it doesn’t fit into either of the above two categories.
Since the set of object instance methods on pandas data structures are generally rich and expressive, we often simply want to invoke, say, a DataFrame function on each group. The name GroupBy should be quite familiar to those who have used a SQL-based tool (or itertools), in which you can write code like:
itertools
SELECT Column1, Column2, mean(Column3), sum(Column4) FROM SomeTable GROUP BY Column1, Column2
We aim to make operations like this natural and easy to express using pandas. We’ll address each area of GroupBy functionality then provide some non-trivial examples / use cases.
See the cookbook for some advanced strategies.
pandas objects can be split on any of their axes. The abstract definition of grouping is to provide a mapping of labels to group names. To create a GroupBy object (more on what the GroupBy object is later), you may do the following:
In [1]: df = pd.DataFrame( ...: [ ...: ("bird", "Falconiformes", 389.0), ...: ("bird", "Psittaciformes", 24.0), ...: ("mammal", "Carnivora", 80.2), ...: ("mammal", "Primates", np.nan), ...: ("mammal", "Carnivora", 58), ...: ], ...: index=["falcon", "parrot", "lion", "monkey", "leopard"], ...: columns=("class", "order", "max_speed"), ...: ) ...: In [2]: df Out[2]: class order max_speed falcon bird Falconiformes 389.0 parrot bird Psittaciformes 24.0 lion mammal Carnivora 80.2 monkey mammal Primates NaN leopard mammal Carnivora 58.0 # default is axis=0 In [3]: grouped = df.groupby("class") In [4]: grouped = df.groupby("order", axis="columns") In [5]: grouped = df.groupby(["class", "order"])
The mapping can be specified many different ways:
A Python function, to be called on each of the axis labels.
A list or NumPy array of the same length as the selected axis.
A dict or Series, providing a label -> group name mapping.
Series
label -> group name
For DataFrame objects, a string indicating either a column name or an index level name to be used to group.
DataFrame
df.groupby('A') is just syntactic sugar for df.groupby(df['A']).
df.groupby('A')
df.groupby(df['A'])
A list of any of the above things.
Collectively we refer to the grouping objects as the keys. For example, consider the following DataFrame:
Note
A string passed to groupby may refer to either a column or an index level. If a string matches both a column name and an index level name, a ValueError will be raised.
groupby
ValueError
In [6]: df = pd.DataFrame( ...: { ...: "A": ["foo", "bar", "foo", "bar", "foo", "bar", "foo", "foo"], ...: "B": ["one", "one", "two", "three", "two", "two", "one", "three"], ...: "C": np.random.randn(8), ...: "D": np.random.randn(8), ...: } ...: ) ...: In [7]: df Out[7]: A B C D 0 foo one 0.469112 -0.861849 1 bar one -0.282863 -2.104569 2 foo two -1.509059 -0.494929 3 bar three -1.135632 1.071804 4 foo two 1.212112 0.721555 5 bar two -0.173215 -0.706771 6 foo one 0.119209 -1.039575 7 foo three -1.044236 0.271860
On a DataFrame, we obtain a GroupBy object by calling groupby(). We could naturally group by either the A or B columns, or both:
groupby()
A
B
In [8]: grouped = df.groupby("A") In [9]: grouped = df.groupby(["A", "B"])
New in version 0.24.
If we also have a MultiIndex on columns A and B, we can group by all but the specified columns
In [10]: df2 = df.set_index(["A", "B"]) In [11]: grouped = df2.groupby(level=df2.index.names.difference(["B"])) In [12]: grouped.sum() Out[12]: C D A bar -1.591710 -1.739537 foo -0.752861 -1.402938
These will split the DataFrame on its index (rows). We could also split by the columns:
In [13]: def get_letter_type(letter): ....: if letter.lower() in 'aeiou': ....: return 'vowel' ....: else: ....: return 'consonant' ....: In [14]: grouped = df.groupby(get_letter_type, axis=1)
pandas Index objects support duplicate values. If a non-unique index is used as the group key in a groupby operation, all values for the same index value will be considered to be in one group and thus the output of aggregation functions will only contain unique index values:
Index
In [15]: lst = [1, 2, 3, 1, 2, 3] In [16]: s = pd.Series([1, 2, 3, 10, 20, 30], lst) In [17]: grouped = s.groupby(level=0) In [18]: grouped.first() Out[18]: 1 1 2 2 3 3 dtype: int64 In [19]: grouped.last() Out[19]: 1 10 2 20 3 30 dtype: int64 In [20]: grouped.sum() Out[20]: 1 11 2 22 3 33 dtype: int64
Note that no splitting occurs until it’s needed. Creating the GroupBy object only verifies that you’ve passed a valid mapping.
Many kinds of complicated data manipulations can be expressed in terms of GroupBy operations (though can’t be guaranteed to be the most efficient). You can get quite creative with the label mapping functions.
By default the group keys are sorted during the groupby operation. You may however pass sort=False for potential speedups:
sort=False
In [21]: df2 = pd.DataFrame({"X": ["B", "B", "A", "A"], "Y": [1, 2, 3, 4]}) In [22]: df2.groupby(["X"]).sum() Out[22]: Y X A 7 B 3 In [23]: df2.groupby(["X"], sort=False).sum() Out[23]: Y X B 3 A 7
Note that groupby will preserve the order in which observations are sorted within each group. For example, the groups created by groupby() below are in the order they appeared in the original DataFrame:
In [24]: df3 = pd.DataFrame({"X": ["A", "B", "A", "B"], "Y": [1, 4, 3, 2]}) In [25]: df3.groupby(["X"]).get_group("A") Out[25]: X Y 0 A 1 2 A 3 In [26]: df3.groupby(["X"]).get_group("B") Out[26]: X Y 1 B 4 3 B 2
New in version 1.1.0.
By default NA values are excluded from group keys during the groupby operation. However, in case you want to include NA values in group keys, you could pass dropna=False to achieve it.
NA
dropna=False
In [27]: df_list = [[1, 2, 3], [1, None, 4], [2, 1, 3], [1, 2, 2]] In [28]: df_dropna = pd.DataFrame(df_list, columns=["a", "b", "c"]) In [29]: df_dropna Out[29]: a b c 0 1 2.0 3 1 1 NaN 4 2 2 1.0 3 3 1 2.0 2
# Default ``dropna`` is set to True, which will exclude NaNs in keys In [30]: df_dropna.groupby(by=["b"], dropna=True).sum() Out[30]: a c b 1.0 2 3 2.0 2 5 # In order to allow NaN in keys, set ``dropna`` to False In [31]: df_dropna.groupby(by=["b"], dropna=False).sum() Out[31]: a c b 1.0 2 3 2.0 2 5 NaN 1 4
The default setting of dropna argument is True which means NA are not included in group keys.
dropna
True
The groups attribute is a dict whose keys are the computed unique groups and corresponding values being the axis labels belonging to each group. In the above example we have:
groups
In [32]: df.groupby("A").groups Out[32]: {'bar': [1, 3, 5], 'foo': [0, 2, 4, 6, 7]} In [33]: df.groupby(get_letter_type, axis=1).groups Out[33]: {'consonant': ['B', 'C', 'D'], 'vowel': ['A']}
Calling the standard Python len function on the GroupBy object just returns the length of the groups dict, so it is largely just a convenience:
len
In [34]: grouped = df.groupby(["A", "B"]) In [35]: grouped.groups Out[35]: {('bar', 'one'): [1], ('bar', 'three'): [3], ('bar', 'two'): [5], ('foo', 'one'): [0, 6], ('foo', 'three'): [7], ('foo', 'two'): [2, 4]} In [36]: len(grouped) Out[36]: 6
GroupBy will tab complete column names (and other attributes):
GroupBy
In [37]: df Out[37]: height weight gender 2000-01-01 42.849980 157.500553 male 2000-01-02 49.607315 177.340407 male 2000-01-03 56.293531 171.524640 male 2000-01-04 48.421077 144.251986 female 2000-01-05 46.556882 152.526206 male 2000-01-06 68.448851 168.272968 female 2000-01-07 70.757698 136.431469 male 2000-01-08 58.909500 176.499753 female 2000-01-09 76.435631 174.094104 female 2000-01-10 45.306120 177.540920 male In [38]: gb = df.groupby("gender")
In [39]: gb.<TAB> # noqa: E225, E999 gb.agg gb.boxplot gb.cummin gb.describe gb.filter gb.get_group gb.height gb.last gb.median gb.ngroups gb.plot gb.rank gb.std gb.transform gb.aggregate gb.count gb.cumprod gb.dtype gb.first gb.groups gb.hist gb.max gb.min gb.nth gb.prod gb.resample gb.sum gb.var gb.apply gb.cummax gb.cumsum gb.fillna gb.gender gb.head gb.indices gb.mean gb.name gb.ohlc gb.quantile gb.size gb.tail gb.weight
With hierarchically-indexed data, it’s quite natural to group by one of the levels of the hierarchy.
Let’s create a Series with a two-level MultiIndex.
MultiIndex
In [40]: arrays = [ ....: ["bar", "bar", "baz", "baz", "foo", "foo", "qux", "qux"], ....: ["one", "two", "one", "two", "one", "two", "one", "two"], ....: ] ....: In [41]: index = pd.MultiIndex.from_arrays(arrays, names=["first", "second"]) In [42]: s = pd.Series(np.random.randn(8), index=index) In [43]: s Out[43]: first second bar one -0.919854 two -0.042379 baz one 1.247642 two -0.009920 foo one 0.290213 two 0.495767 qux one 0.362949 two 1.548106 dtype: float64
We can then group by one of the levels in s.
s
In [44]: grouped = s.groupby(level=0) In [45]: grouped.sum() Out[45]: first bar -0.962232 baz 1.237723 foo 0.785980 qux 1.911055 dtype: float64
If the MultiIndex has names specified, these can be passed instead of the level number:
In [46]: s.groupby(level="second").sum() Out[46]: second one 0.980950 two 1.991575 dtype: float64
The aggregation functions such as sum will take the level parameter directly. Additionally, the resulting index will be named according to the chosen level:
sum
In [47]: s.sum(level="second") Out[47]: second one 0.980950 two 1.991575 dtype: float64
Grouping with multiple levels is supported.
In [48]: s Out[48]: first second third bar doo one -1.131345 two -0.089329 baz bee one 0.337863 two -0.945867 foo bop one -0.932132 two 1.956030 qux bop one 0.017587 two -0.016692 dtype: float64 In [49]: s.groupby(level=["first", "second"]).sum() Out[49]: first second bar doo -1.220674 baz bee -0.608004 foo bop 1.023898 qux bop 0.000895 dtype: float64
Index level names may be supplied as keys.
In [50]: s.groupby(["first", "second"]).sum() Out[50]: first second bar doo -1.220674 baz bee -0.608004 foo bop 1.023898 qux bop 0.000895 dtype: float64
More on the sum function and aggregation later.
A DataFrame may be grouped by a combination of columns and index levels by specifying the column names as strings and the index levels as pd.Grouper objects.
pd.Grouper
In [51]: arrays = [ ....: ["bar", "bar", "baz", "baz", "foo", "foo", "qux", "qux"], ....: ["one", "two", "one", "two", "one", "two", "one", "two"], ....: ] ....: In [52]: index = pd.MultiIndex.from_arrays(arrays, names=["first", "second"]) In [53]: df = pd.DataFrame({"A": [1, 1, 1, 1, 2, 2, 3, 3], "B": np.arange(8)}, index=index) In [54]: df Out[54]: A B first second bar one 1 0 two 1 1 baz one 1 2 two 1 3 foo one 2 4 two 2 5 qux one 3 6 two 3 7
The following example groups df by the second index level and the A column.
df
second
In [55]: df.groupby([pd.Grouper(level=1), "A"]).sum() Out[55]: B second A one 1 2 2 4 3 6 two 1 4 2 5 3 7
Index levels may also be specified by name.
In [56]: df.groupby([pd.Grouper(level="second"), "A"]).sum() Out[56]: B second A one 1 2 2 4 3 6 two 1 4 2 5 3 7
Index level names may be specified as keys directly to groupby.
In [57]: df.groupby(["second", "A"]).sum() Out[57]: B second A one 1 2 2 4 3 6 two 1 4 2 5 3 7
Once you have created the GroupBy object from a DataFrame, you might want to do something different for each of the columns. Thus, using [] similar to getting a column from a DataFrame, you can do:
[]
In [58]: grouped = df.groupby(["A"]) In [59]: grouped_C = grouped["C"] In [60]: grouped_D = grouped["D"]
This is mainly syntactic sugar for the alternative and much more verbose:
In [61]: df["C"].groupby(df["A"]) Out[61]: <pandas.core.groupby.generic.SeriesGroupBy object at 0x7fa941fbb370>
Additionally this method avoids recomputing the internal grouping information derived from the passed key.
With the GroupBy object in hand, iterating through the grouped data is very natural and functions similarly to itertools.groupby():
itertools.groupby()
In [62]: grouped = df.groupby('A') In [63]: for name, group in grouped: ....: print(name) ....: print(group) ....: bar A B C D 1 bar one 0.254161 1.511763 3 bar three 0.215897 -0.990582 5 bar two -0.077118 1.211526 foo A B C D 0 foo one -0.575247 1.346061 2 foo two -1.143704 1.627081 4 foo two 1.193555 -0.441652 6 foo one -0.408530 0.268520 7 foo three -0.862495 0.024580
In the case of grouping by multiple keys, the group name will be a tuple:
In [64]: for name, group in df.groupby(['A', 'B']): ....: print(name) ....: print(group) ....: ('bar', 'one') A B C D 1 bar one 0.254161 1.511763 ('bar', 'three') A B C D 3 bar three 0.215897 -0.990582 ('bar', 'two') A B C D 5 bar two -0.077118 1.211526 ('foo', 'one') A B C D 0 foo one -0.575247 1.346061 6 foo one -0.408530 0.268520 ('foo', 'three') A B C D 7 foo three -0.862495 0.02458 ('foo', 'two') A B C D 2 foo two -1.143704 1.627081 4 foo two 1.193555 -0.441652
See Iterating through groups.
A single group can be selected using get_group():
get_group()
In [65]: grouped.get_group("bar") Out[65]: A B C D 1 bar one 0.254161 1.511763 3 bar three 0.215897 -0.990582 5 bar two -0.077118 1.211526
Or for an object grouped on multiple columns:
In [66]: df.groupby(["A", "B"]).get_group(("bar", "one")) Out[66]: A B C D 1 bar one 0.254161 1.511763
Once the GroupBy object has been created, several methods are available to perform a computation on the grouped data. These operations are similar to the aggregating API, window API, and resample API.
An obvious one is aggregation via the aggregate() or equivalently agg() method:
aggregate()
agg()
In [67]: grouped = df.groupby("A") In [68]: grouped.aggregate(np.sum) Out[68]: C D A bar 0.392940 1.732707 foo -1.796421 2.824590 In [69]: grouped = df.groupby(["A", "B"]) In [70]: grouped.aggregate(np.sum) Out[70]: C D A B bar one 0.254161 1.511763 three 0.215897 -0.990582 two -0.077118 1.211526 foo one -0.983776 1.614581 three -0.862495 0.024580 two 0.049851 1.185429
As you can see, the result of the aggregation will have the group names as the new index along the grouped axis. In the case of multiple keys, the result is a MultiIndex by default, though this can be changed by using the as_index option:
as_index
In [71]: grouped = df.groupby(["A", "B"], as_index=False) In [72]: grouped.aggregate(np.sum) Out[72]: A B C D 0 bar one 0.254161 1.511763 1 bar three 0.215897 -0.990582 2 bar two -0.077118 1.211526 3 foo one -0.983776 1.614581 4 foo three -0.862495 0.024580 5 foo two 0.049851 1.185429 In [73]: df.groupby("A", as_index=False).sum() Out[73]: A C D 0 bar 0.392940 1.732707 1 foo -1.796421 2.824590
Note that you could use the reset_index DataFrame function to achieve the same result as the column names are stored in the resulting MultiIndex:
reset_index
In [74]: df.groupby(["A", "B"]).sum().reset_index() Out[74]: A B C D 0 bar one 0.254161 1.511763 1 bar three 0.215897 -0.990582 2 bar two -0.077118 1.211526 3 foo one -0.983776 1.614581 4 foo three -0.862495 0.024580 5 foo two 0.049851 1.185429
Another simple aggregation example is to compute the size of each group. This is included in GroupBy as the size method. It returns a Series whose index are the group names and whose values are the sizes of each group.
size
In [75]: grouped.size() Out[75]: A B size 0 bar one 1 1 bar three 1 2 bar two 1 3 foo one 2 4 foo three 1 5 foo two 2
In [76]: grouped.describe() Out[76]: C ... D count mean std min 25% 50% ... std min 25% 50% 75% max 0 1.0 0.254161 NaN 0.254161 0.254161 0.254161 ... NaN 1.511763 1.511763 1.511763 1.511763 1.511763 1 1.0 0.215897 NaN 0.215897 0.215897 0.215897 ... NaN -0.990582 -0.990582 -0.990582 -0.990582 -0.990582 2 1.0 -0.077118 NaN -0.077118 -0.077118 -0.077118 ... NaN 1.211526 1.211526 1.211526 1.211526 1.211526 3 2.0 -0.491888 0.117887 -0.575247 -0.533567 -0.491888 ... 0.761937 0.268520 0.537905 0.807291 1.076676 1.346061 4 1.0 -0.862495 NaN -0.862495 -0.862495 -0.862495 ... NaN 0.024580 0.024580 0.024580 0.024580 0.024580 5 2.0 0.024925 1.652692 -1.143704 -0.559389 0.024925 ... 1.462816 -0.441652 0.075531 0.592714 1.109898 1.627081 [6 rows x 16 columns]
Another aggregation example is to compute the number of unique values of each group. This is similar to the value_counts function, except that it only counts unique values.
value_counts
In [77]: ll = [['foo', 1], ['foo', 2], ['foo', 2], ['bar', 1], ['bar', 1]] In [78]: df4 = pd.DataFrame(ll, columns=["A", "B"]) In [79]: df4 Out[79]: A B 0 foo 1 1 foo 2 2 foo 2 3 bar 1 4 bar 1 In [80]: df4.groupby("A")["B"].nunique() Out[80]: A bar 1 foo 2 Name: B, dtype: int64
Aggregation functions will not return the groups that you are aggregating over if they are named columns, when as_index=True, the default. The grouped columns will be the indices of the returned object.
as_index=True
Passing as_index=False will return the groups that you are aggregating over, if they are named columns.
as_index=False
Aggregating functions are the ones that reduce the dimension of the returned objects. Some common aggregating functions are tabulated below:
Function
Description
mean()
Compute mean of groups
sum()
Compute sum of group values
size()
Compute group sizes
count()
Compute count of group
std()
Standard deviation of groups
var()
Compute variance of groups
sem()
Standard error of the mean of groups
describe()
Generates descriptive statistics
first()
Compute first of group values
last()
Compute last of group values
nth()
Take nth value, or a subset if n is a list
min()
Compute min of group values
max()
Compute max of group values
The aggregating functions above will exclude NA values. Any function which reduces a Series to a scalar value is an aggregation function and will work, a trivial example is df.groupby('A').agg(lambda ser: 1). Note that nth() can act as a reducer or a filter, see here.
df.groupby('A').agg(lambda ser: 1)
With grouped Series you can also pass a list or dict of functions to do aggregation with, outputting a DataFrame:
In [81]: grouped = df.groupby("A") In [82]: grouped["C"].agg([np.sum, np.mean, np.std]) Out[82]: sum mean std A bar 0.392940 0.130980 0.181231 foo -1.796421 -0.359284 0.912265
On a grouped DataFrame, you can pass a list of functions to apply to each column, which produces an aggregated result with a hierarchical index:
In [83]: grouped.agg([np.sum, np.mean, np.std]) Out[83]: C D sum mean std sum mean std A bar 0.392940 0.130980 0.181231 1.732707 0.577569 1.366330 foo -1.796421 -0.359284 0.912265 2.824590 0.564918 0.884785
The resulting aggregations are named for the functions themselves. If you need to rename, then you can add in a chained operation for a Series like this:
In [84]: ( ....: grouped["C"] ....: .agg([np.sum, np.mean, np.std]) ....: .rename(columns={"sum": "foo", "mean": "bar", "std": "baz"}) ....: ) ....: Out[84]: foo bar baz A bar 0.392940 0.130980 0.181231 foo -1.796421 -0.359284 0.912265
For a grouped DataFrame, you can rename in a similar manner:
In [85]: ( ....: grouped.agg([np.sum, np.mean, np.std]).rename( ....: columns={"sum": "foo", "mean": "bar", "std": "baz"} ....: ) ....: ) ....: Out[85]: C D foo bar baz foo bar baz A bar 0.392940 0.130980 0.181231 1.732707 0.577569 1.366330 foo -1.796421 -0.359284 0.912265 2.824590 0.564918 0.884785
In general, the output column names should be unique. You can’t apply the same function (or two functions with the same name) to the same column.
In [86]: grouped["C"].agg(["sum", "sum"]) Out[86]: sum sum A bar 0.392940 0.392940 foo -1.796421 -1.796421
pandas does allow you to provide multiple lambdas. In this case, pandas will mangle the name of the (nameless) lambda functions, appending _<i> to each subsequent lambda.
_<i>
In [87]: grouped["C"].agg([lambda x: x.max() - x.min(), lambda x: x.median() - x.mean()]) Out[87]: <lambda_0> <lambda_1> A bar 0.331279 0.084917 foo 2.337259 -0.215962
New in version 0.25.0.
To support column-specific aggregation with control over the output column names, pandas accepts the special syntax in GroupBy.agg(), known as “named aggregation”, where
GroupBy.agg()
The keywords are the output column names
The values are tuples whose first element is the column to select and the second element is the aggregation to apply to that column. pandas provides the pandas.NamedAgg namedtuple with the fields ['column', 'aggfunc'] to make it clearer what the arguments are. As usual, the aggregation can be a callable or a string alias.
pandas.NamedAgg
['column', 'aggfunc']
In [88]: animals = pd.DataFrame( ....: { ....: "kind": ["cat", "dog", "cat", "dog"], ....: "height": [9.1, 6.0, 9.5, 34.0], ....: "weight": [7.9, 7.5, 9.9, 198.0], ....: } ....: ) ....: In [89]: animals Out[89]: kind height weight 0 cat 9.1 7.9 1 dog 6.0 7.5 2 cat 9.5 9.9 3 dog 34.0 198.0 In [90]: animals.groupby("kind").agg( ....: min_height=pd.NamedAgg(column="height", aggfunc="min"), ....: max_height=pd.NamedAgg(column="height", aggfunc="max"), ....: average_weight=pd.NamedAgg(column="weight", aggfunc=np.mean), ....: ) ....: Out[90]: min_height max_height average_weight kind cat 9.1 9.5 8.90 dog 6.0 34.0 102.75
pandas.NamedAgg is just a namedtuple. Plain tuples are allowed as well.
namedtuple
In [91]: animals.groupby("kind").agg( ....: min_height=("height", "min"), ....: max_height=("height", "max"), ....: average_weight=("weight", np.mean), ....: ) ....: Out[91]: min_height max_height average_weight kind cat 9.1 9.5 8.90 dog 6.0 34.0 102.75
If your desired output column names are not valid Python keywords, construct a dictionary and unpack the keyword arguments
In [92]: animals.groupby("kind").agg( ....: **{ ....: "total weight": pd.NamedAgg(column="weight", aggfunc=sum) ....: } ....: ) ....: Out[92]: total weight kind cat 17.8 dog 205.5
Additional keyword arguments are not passed through to the aggregation functions. Only pairs of (column, aggfunc) should be passed as **kwargs. If your aggregation functions requires additional arguments, partially apply them with functools.partial().
(column, aggfunc)
**kwargs
functools.partial()
For Python 3.5 and earlier, the order of **kwargs in a functions was not preserved. This means that the output column ordering would not be consistent. To ensure consistent ordering, the keys (and so output columns) will always be sorted for Python 3.5.
Named aggregation is also valid for Series groupby aggregations. In this case there’s no column selection, so the values are just the functions.
In [93]: animals.groupby("kind").height.agg( ....: min_height="min", ....: max_height="max", ....: ) ....: Out[93]: min_height max_height kind cat 9.1 9.5 dog 6.0 34.0
By passing a dict to aggregate you can apply a different aggregation to the columns of a DataFrame:
aggregate
In [94]: grouped.agg({"C": np.sum, "D": lambda x: np.std(x, ddof=1)}) Out[94]: C D A bar 0.392940 1.366330 foo -1.796421 0.884785
The function names can also be strings. In order for a string to be valid it must be either implemented on GroupBy or available via dispatching:
In [95]: grouped.agg({"C": "sum", "D": "std"}) Out[95]: C D A bar 0.392940 1.366330 foo -1.796421 0.884785
Some common aggregations, currently only sum, mean, std, and sem, have optimized Cython implementations:
mean
std
sem
In [96]: df.groupby("A").sum() Out[96]: C D A bar 0.392940 1.732707 foo -1.796421 2.824590 In [97]: df.groupby(["A", "B"]).mean() Out[97]: C D A B bar one 0.254161 1.511763 three 0.215897 -0.990582 two -0.077118 1.211526 foo one -0.491888 0.807291 three -0.862495 0.024580 two 0.024925 0.592714
Of course sum and mean are implemented on pandas objects, so the above code would work even without the special versions via dispatching (see below).
The transform method returns an object that is indexed the same (same size) as the one being grouped. The transform function must:
transform
Return a result that is either the same size as the group chunk or broadcastable to the size of the group chunk (e.g., a scalar, grouped.transform(lambda x: x.iloc[-1])).
grouped.transform(lambda x: x.iloc[-1])
Operate column-by-column on the group chunk. The transform is applied to the first group chunk using chunk.apply.
Not perform in-place operations on the group chunk. Group chunks should be treated as immutable, and changes to a group chunk may produce unexpected results. For example, when using fillna, inplace must be False (grouped.transform(lambda x: x.fillna(inplace=False))).
fillna
inplace
False
grouped.transform(lambda x: x.fillna(inplace=False))
(Optionally) operates on the entire group chunk. If this is supported, a fast path is used starting from the second chunk.
For example, suppose we wished to standardize the data within each group:
In [98]: index = pd.date_range("10/1/1999", periods=1100) In [99]: ts = pd.Series(np.random.normal(0.5, 2, 1100), index) In [100]: ts = ts.rolling(window=100, min_periods=100).mean().dropna() In [101]: ts.head() Out[101]: 2000-01-08 0.779333 2000-01-09 0.778852 2000-01-10 0.786476 2000-01-11 0.782797 2000-01-12 0.798110 Freq: D, dtype: float64 In [102]: ts.tail() Out[102]: 2002-09-30 0.660294 2002-10-01 0.631095 2002-10-02 0.673601 2002-10-03 0.709213 2002-10-04 0.719369 Freq: D, dtype: float64 In [103]: transformed = ts.groupby(lambda x: x.year).transform( .....: lambda x: (x - x.mean()) / x.std() .....: ) .....:
We would expect the result to now have mean 0 and standard deviation 1 within each group, which we can easily check:
# Original Data In [104]: grouped = ts.groupby(lambda x: x.year) In [105]: grouped.mean() Out[105]: 2000 0.442441 2001 0.526246 2002 0.459365 dtype: float64 In [106]: grouped.std() Out[106]: 2000 0.131752 2001 0.210945 2002 0.128753 dtype: float64 # Transformed Data In [107]: grouped_trans = transformed.groupby(lambda x: x.year) In [108]: grouped_trans.mean() Out[108]: 2000 1.167126e-15 2001 2.190637e-15 2002 1.088580e-15 dtype: float64 In [109]: grouped_trans.std() Out[109]: 2000 1.0 2001 1.0 2002 1.0 dtype: float64
We can also visually compare the original and transformed data sets.
In [110]: compare = pd.DataFrame({"Original": ts, "Transformed": transformed}) In [111]: compare.plot() Out[111]: <AxesSubplot:>
Transformation functions that have lower dimension outputs are broadcast to match the shape of the input array.
In [112]: ts.groupby(lambda x: x.year).transform(lambda x: x.max() - x.min()) Out[112]: 2000-01-08 0.623893 2000-01-09 0.623893 2000-01-10 0.623893 2000-01-11 0.623893 2000-01-12 0.623893 ... 2002-09-30 0.558275 2002-10-01 0.558275 2002-10-02 0.558275 2002-10-03 0.558275 2002-10-04 0.558275 Freq: D, Length: 1001, dtype: float64
Alternatively, the built-in methods could be used to produce the same outputs.
In [113]: max = ts.groupby(lambda x: x.year).transform("max") In [114]: min = ts.groupby(lambda x: x.year).transform("min") In [115]: max - min Out[115]: 2000-01-08 0.623893 2000-01-09 0.623893 2000-01-10 0.623893 2000-01-11 0.623893 2000-01-12 0.623893 ... 2002-09-30 0.558275 2002-10-01 0.558275 2002-10-02 0.558275 2002-10-03 0.558275 2002-10-04 0.558275 Freq: D, Length: 1001, dtype: float64
Another common data transform is to replace missing data with the group mean.
In [116]: data_df Out[116]: A B C 0 1.539708 -1.166480 0.533026 1 1.302092 -0.505754 NaN 2 -0.371983 1.104803 -0.651520 3 -1.309622 1.118697 -1.161657 4 -1.924296 0.396437 0.812436 .. ... ... ... 995 -0.093110 0.683847 -0.774753 996 -0.185043 1.438572 NaN 997 -0.394469 -0.642343 0.011374 998 -1.174126 1.857148 NaN 999 0.234564 0.517098 0.393534 [1000 rows x 3 columns] In [117]: countries = np.array(["US", "UK", "GR", "JP"]) In [118]: key = countries[np.random.randint(0, 4, 1000)] In [119]: grouped = data_df.groupby(key) # Non-NA count in each group In [120]: grouped.count() Out[120]: A B C GR 209 217 189 JP 240 255 217 UK 216 231 193 US 239 250 217 In [121]: transformed = grouped.transform(lambda x: x.fillna(x.mean()))
We can verify that the group means have not changed in the transformed data and that the transformed data contains no NAs.
In [122]: grouped_trans = transformed.groupby(key) In [123]: grouped.mean() # original group means Out[123]: A B C GR -0.098371 -0.015420 0.068053 JP 0.069025 0.023100 -0.077324 UK 0.034069 -0.052580 -0.116525 US 0.058664 -0.020399 0.028603 In [124]: grouped_trans.mean() # transformation did not change group means Out[124]: A B C GR -0.098371 -0.015420 0.068053 JP 0.069025 0.023100 -0.077324 UK 0.034069 -0.052580 -0.116525 US 0.058664 -0.020399 0.028603 In [125]: grouped.count() # original has some missing data points Out[125]: A B C GR 209 217 189 JP 240 255 217 UK 216 231 193 US 239 250 217 In [126]: grouped_trans.count() # counts after transformation Out[126]: A B C GR 228 228 228 JP 267 267 267 UK 247 247 247 US 258 258 258 In [127]: grouped_trans.size() # Verify non-NA count equals group size Out[127]: GR 228 JP 267 UK 247 US 258 dtype: int64
Some functions will automatically transform the input when applied to a GroupBy object, but returning an object of the same shape as the original. Passing as_index=False will not affect these transformation methods.
For example: fillna, ffill, bfill, shift..
fillna, ffill, bfill, shift.
In [128]: grouped.ffill() Out[128]: A B C 0 1.539708 -1.166480 0.533026 1 1.302092 -0.505754 0.533026 2 -0.371983 1.104803 -0.651520 3 -1.309622 1.118697 -1.161657 4 -1.924296 0.396437 0.812436 .. ... ... ... 995 -0.093110 0.683847 -0.774753 996 -0.185043 1.438572 -0.774753 997 -0.394469 -0.642343 0.011374 998 -1.174126 1.857148 -0.774753 999 0.234564 0.517098 0.393534 [1000 rows x 3 columns]
It is possible to use resample(), expanding() and rolling() as methods on groupbys.
resample()
expanding()
rolling()
The example below will apply the rolling() method on the samples of the column B based on the groups of column A.
In [129]: df_re = pd.DataFrame({"A": [1] * 10 + [5] * 10, "B": np.arange(20)}) In [130]: df_re Out[130]: A B 0 1 0 1 1 1 2 1 2 3 1 3 4 1 4 .. .. .. 15 5 15 16 5 16 17 5 17 18 5 18 19 5 19 [20 rows x 2 columns] In [131]: df_re.groupby("A").rolling(4).B.mean() Out[131]: A 1 0 NaN 1 NaN 2 NaN 3 1.5 4 2.5 ... 5 15 13.5 16 14.5 17 15.5 18 16.5 19 17.5 Name: B, Length: 20, dtype: float64
The expanding() method will accumulate a given operation (sum() in the example) for all the members of each particular group.
In [132]: df_re.groupby("A").expanding().sum() Out[132]: A B A 1 0 1.0 0.0 1 2.0 1.0 2 3.0 3.0 3 4.0 6.0 4 5.0 10.0 ... ... ... 5 15 30.0 75.0 16 35.0 91.0 17 40.0 108.0 18 45.0 126.0 19 50.0 145.0 [20 rows x 2 columns]
Suppose you want to use the resample() method to get a daily frequency in each group of your dataframe and wish to complete the missing values with the ffill() method.
ffill()
In [133]: df_re = pd.DataFrame( .....: { .....: "date": pd.date_range(start="2016-01-01", periods=4, freq="W"), .....: "group": [1, 1, 2, 2], .....: "val": [5, 6, 7, 8], .....: } .....: ).set_index("date") .....: In [134]: df_re Out[134]: group val date 2016-01-03 1 5 2016-01-10 1 6 2016-01-17 2 7 2016-01-24 2 8 In [135]: df_re.groupby("group").resample("1D").ffill() Out[135]: group val group date 1 2016-01-03 1 5 2016-01-04 1 5 2016-01-05 1 5 2016-01-06 1 5 2016-01-07 1 5 ... ... ... 2 2016-01-20 2 7 2016-01-21 2 7 2016-01-22 2 7 2016-01-23 2 7 2016-01-24 2 8 [16 rows x 2 columns]
The filter method returns a subset of the original object. Suppose we want to take only elements that belong to groups with a group sum greater than 2.
filter
In [136]: sf = pd.Series([1, 1, 2, 3, 3, 3]) In [137]: sf.groupby(sf).filter(lambda x: x.sum() > 2) Out[137]: 3 3 4 3 5 3 dtype: int64
The argument of filter must be a function that, applied to the group as a whole, returns True or False.
Another useful operation is filtering out elements that belong to groups with only a couple members.
In [138]: dff = pd.DataFrame({"A": np.arange(8), "B": list("aabbbbcc")}) In [139]: dff.groupby("B").filter(lambda x: len(x) > 2) Out[139]: A B 2 2 b 3 3 b 4 4 b 5 5 b
Alternatively, instead of dropping the offending groups, we can return a like-indexed objects where the groups that do not pass the filter are filled with NaNs.
In [140]: dff.groupby("B").filter(lambda x: len(x) > 2, dropna=False) Out[140]: A B 0 NaN NaN 1 NaN NaN 2 2.0 b 3 3.0 b 4 4.0 b 5 5.0 b 6 NaN NaN 7 NaN NaN
For DataFrames with multiple columns, filters should explicitly specify a column as the filter criterion.
In [141]: dff["C"] = np.arange(8) In [142]: dff.groupby("B").filter(lambda x: len(x["C"]) > 2) Out[142]: A B C 2 2 b 2 3 3 b 3 4 4 b 4 5 5 b 5
Some functions when applied to a groupby object will act as a filter on the input, returning a reduced shape of the original (and potentially eliminating groups), but with the index unchanged. Passing as_index=False will not affect these transformation methods.
For example: head, tail.
head, tail
In [143]: dff.groupby("B").head(2) Out[143]: A B C 0 0 a 0 1 1 a 1 2 2 b 2 3 3 b 3 6 6 c 6 7 7 c 7
When doing an aggregation or transformation, you might just want to call an instance method on each data group. This is pretty easy to do by passing lambda functions:
In [144]: grouped = df.groupby("A") In [145]: grouped.agg(lambda x: x.std()) Out[145]: C D A bar 0.181231 1.366330 foo 0.912265 0.884785
But, it’s rather verbose and can be untidy if you need to pass additional arguments. Using a bit of metaprogramming cleverness, GroupBy now has the ability to “dispatch” method calls to the groups:
In [146]: grouped.std() Out[146]: C D A bar 0.181231 1.366330 foo 0.912265 0.884785
What is actually happening here is that a function wrapper is being generated. When invoked, it takes any passed arguments and invokes the function with any arguments on each group (in the above example, the std function). The results are then combined together much in the style of agg and transform (it actually uses apply to infer the gluing, documented next). This enables some operations to be carried out rather succinctly:
agg
apply
In [147]: tsdf = pd.DataFrame( .....: np.random.randn(1000, 3), .....: index=pd.date_range("1/1/2000", periods=1000), .....: columns=["A", "B", "C"], .....: ) .....: In [148]: tsdf.iloc[::2] = np.nan In [149]: grouped = tsdf.groupby(lambda x: x.year) In [150]: grouped.fillna(method="pad") Out[150]: A B C 2000-01-01 NaN NaN NaN 2000-01-02 -0.353501 -0.080957 -0.876864 2000-01-03 -0.353501 -0.080957 -0.876864 2000-01-04 0.050976 0.044273 -0.559849 2000-01-05 0.050976 0.044273 -0.559849 ... ... ... ... 2002-09-22 0.005011 0.053897 -1.026922 2002-09-23 0.005011 0.053897 -1.026922 2002-09-24 -0.456542 -1.849051 1.559856 2002-09-25 -0.456542 -1.849051 1.559856 2002-09-26 1.123162 0.354660 1.128135 [1000 rows x 3 columns]
In this example, we chopped the collection of time series into yearly chunks then independently called fillna on the groups.
The nlargest and nsmallest methods work on Series style groupbys:
nlargest
nsmallest
In [151]: s = pd.Series([9, 8, 7, 5, 19, 1, 4.2, 3.3]) In [152]: g = pd.Series(list("abababab")) In [153]: gb = s.groupby(g) In [154]: gb.nlargest(3) Out[154]: a 4 19.0 0 9.0 2 7.0 b 1 8.0 3 5.0 7 3.3 dtype: float64 In [155]: gb.nsmallest(3) Out[155]: a 6 4.2 2 7.0 0 9.0 b 5 1.0 7 3.3 3 5.0 dtype: float64
Some operations on the grouped data might not fit into either the aggregate or transform categories. Or, you may simply want GroupBy to infer how to combine the results. For these, use the apply function, which can be substituted for both aggregate and transform in many standard use cases. However, apply can handle some exceptional use cases, for example:
In [156]: df Out[156]: A B C D 0 foo one -0.575247 1.346061 1 bar one 0.254161 1.511763 2 foo two -1.143704 1.627081 3 bar three 0.215897 -0.990582 4 foo two 1.193555 -0.441652 5 bar two -0.077118 1.211526 6 foo one -0.408530 0.268520 7 foo three -0.862495 0.024580 In [157]: grouped = df.groupby("A") # could also just call .describe() In [158]: grouped["C"].apply(lambda x: x.describe()) Out[158]: A bar count 3.000000 mean 0.130980 std 0.181231 min -0.077118 25% 0.069390 ... foo min -1.143704 25% -0.862495 50% -0.575247 75% -0.408530 max 1.193555 Name: C, Length: 16, dtype: float64
The dimension of the returned result can also change:
In [159]: grouped = df.groupby('A')['C'] In [160]: def f(group): .....: return pd.DataFrame({'original': group, .....: 'demeaned': group - group.mean()}) .....: In [161]: grouped.apply(f) Out[161]: original demeaned 0 -0.575247 -0.215962 1 0.254161 0.123181 2 -1.143704 -0.784420 3 0.215897 0.084917 4 1.193555 1.552839 5 -0.077118 -0.208098 6 -0.408530 -0.049245 7 -0.862495 -0.503211
apply on a Series can operate on a returned value from the applied function, that is itself a series, and possibly upcast the result to a DataFrame:
In [162]: def f(x): .....: return pd.Series([x, x ** 2], index=["x", "x^2"]) .....: In [163]: s = pd.Series(np.random.rand(5)) In [164]: s Out[164]: 0 0.321438 1 0.493496 2 0.139505 3 0.910103 4 0.194158 dtype: float64 In [165]: s.apply(f) Out[165]: x x^2 0 0.321438 0.103323 1 0.493496 0.243538 2 0.139505 0.019462 3 0.910103 0.828287 4 0.194158 0.037697
apply can act as a reducer, transformer, or filter function, depending on exactly what is passed to it. So depending on the path taken, and exactly what you are grouping. Thus the grouped columns(s) may be included in the output as well as set the indices.
New in version 1.1.
If Numba is installed as an optional dependency, the transform and aggregate methods support engine='numba' and engine_kwargs arguments. The engine_kwargs argument is a dictionary of keyword arguments that will be passed into the numba.jit decorator. These keyword arguments will be applied to the passed function. Currently only nogil, nopython, and parallel are supported, and their default values are set to False, True and False respectively.
engine='numba'
engine_kwargs
nogil
nopython
parallel
The function signature must start with values, index exactly as the data belonging to each group will be passed into values, and the group index will be passed into index.
values, index
values
index
Warning
When using engine='numba', there will be no “fall back” behavior internally. The group data and group index will be passed as NumPy arrays to the JITed user defined function, and no alternative execution attempts will be tried.
In terms of performance, the first time a function is run using the Numba engine will be slow as Numba will have some function compilation overhead. However, the compiled functions are cached, and subsequent calls will be fast. In general, the Numba engine is performant with a larger amount of data points (e.g. 1+ million).
In [1]: N = 10 ** 3 In [2]: data = {0: [str(i) for i in range(100)] * N, 1: list(range(100)) * N} In [3]: df = pd.DataFrame(data, columns=[0, 1]) In [4]: def f_numba(values, index): ...: total = 0 ...: for i, value in enumerate(values): ...: if i % 2: ...: total += value + 5 ...: else: ...: total += value * 2 ...: return total ...: In [5]: def f_cython(values): ...: total = 0 ...: for i, value in enumerate(values): ...: if i % 2: ...: total += value + 5 ...: else: ...: total += value * 2 ...: return total ...: In [6]: groupby = df.groupby(0) # Run the first time, compilation time will affect performance In [7]: %timeit -r 1 -n 1 groupby.aggregate(f_numba, engine='numba') # noqa: E225 2.14 s ± 0 ns per loop (mean ± std. dev. of 1 run, 1 loop each) # Function is cached and performance will improve In [8]: %timeit groupby.aggregate(f_numba, engine='numba') 4.93 ms ± 32.3 µs per loop (mean ± std. dev. of 7 runs, 100 loops each) In [9]: %timeit groupby.aggregate(f_cython, engine='cython') 18.6 ms ± 84.8 µs per loop (mean ± std. dev. of 7 runs, 100 loops each)
Again consider the example DataFrame we’ve been looking at:
In [166]: df Out[166]: A B C D 0 foo one -0.575247 1.346061 1 bar one 0.254161 1.511763 2 foo two -1.143704 1.627081 3 bar three 0.215897 -0.990582 4 foo two 1.193555 -0.441652 5 bar two -0.077118 1.211526 6 foo one -0.408530 0.268520 7 foo three -0.862495 0.024580
Suppose we wish to compute the standard deviation grouped by the A column. There is a slight problem, namely that we don’t care about the data in column B. We refer to this as a “nuisance” column. If the passed aggregation function can’t be applied to some columns, the troublesome columns will be (silently) dropped. Thus, this does not pose any problems:
In [167]: df.groupby("A").std() Out[167]: C D A bar 0.181231 1.366330 foo 0.912265 0.884785
Note that df.groupby('A').colname.std(). is more efficient than df.groupby('A').std().colname, so if the result of an aggregation function is only interesting over one column (here colname), it may be filtered before applying the aggregation function.
df.groupby('A').colname.std().
df.groupby('A').std().colname
colname
Any object column, also if it contains numerical values such as Decimal objects, is considered as a “nuisance” columns. They are excluded from aggregate functions automatically in groupby.
Decimal
If you do wish to include decimal or object columns in an aggregation with other non-nuisance data types, you must do so explicitly.
In [168]: from decimal import Decimal In [169]: df_dec = pd.DataFrame( .....: { .....: "id": [1, 2, 1, 2], .....: "int_column": [1, 2, 3, 4], .....: "dec_column": [ .....: Decimal("0.50"), .....: Decimal("0.15"), .....: Decimal("0.25"), .....: Decimal("0.40"), .....: ], .....: } .....: ) .....: # Decimal columns can be sum'd explicitly by themselves... In [170]: df_dec.groupby(["id"])[["dec_column"]].sum() Out[170]: dec_column id 1 0.75 2 0.55 # ...but cannot be combined with standard data types or they will be excluded In [171]: df_dec.groupby(["id"])[["int_column", "dec_column"]].sum() Out[171]: int_column id 1 4 2 6 # Use .agg function to aggregate over standard and "nuisance" data types # at the same time In [172]: df_dec.groupby(["id"]).agg({"int_column": "sum", "dec_column": "sum"}) Out[172]: int_column dec_column id 1 4 0.75 2 6 0.55
When using a Categorical grouper (as a single grouper, or as part of multiple groupers), the observed keyword controls whether to return a cartesian product of all possible groupers values (observed=False) or only those that are observed groupers (observed=True).
Categorical
observed
observed=False
observed=True
Show all values:
In [173]: pd.Series([1, 1, 1]).groupby( .....: pd.Categorical(["a", "a", "a"], categories=["a", "b"]), observed=False .....: ).count() .....: Out[173]: a 3 b 0 dtype: int64
Show only the observed values:
In [174]: pd.Series([1, 1, 1]).groupby( .....: pd.Categorical(["a", "a", "a"], categories=["a", "b"]), observed=True .....: ).count() .....: Out[174]: a 3 dtype: int64
The returned dtype of the grouped will always include all of the categories that were grouped.
In [175]: s = ( .....: pd.Series([1, 1, 1]) .....: .groupby(pd.Categorical(["a", "a", "a"], categories=["a", "b"]), observed=False) .....: .count() .....: ) .....: In [176]: s.index.dtype Out[176]: CategoricalDtype(categories=['a', 'b'], ordered=False)
If there are any NaN or NaT values in the grouping key, these will be automatically excluded. In other words, there will never be an “NA group” or “NaT group”. This was not the case in older versions of pandas, but users were generally discarding the NA group anyway (and supporting it was an implementation headache).
Categorical variables represented as instance of pandas’s Categorical class can be used as group keys. If so, the order of the levels will be preserved:
In [177]: data = pd.Series(np.random.randn(100)) In [178]: factor = pd.qcut(data, [0, 0.25, 0.5, 0.75, 1.0]) In [179]: data.groupby(factor).mean() Out[179]: (-2.645, -0.523] -1.362896 (-0.523, 0.0296] -0.260266 (0.0296, 0.654] 0.361802 (0.654, 2.21] 1.073801 dtype: float64
You may need to specify a bit more data to properly group. You can use the pd.Grouper to provide this local control.
In [180]: import datetime In [181]: df = pd.DataFrame( .....: { .....: "Branch": "A A A A A A A B".split(), .....: "Buyer": "Carl Mark Carl Carl Joe Joe Joe Carl".split(), .....: "Quantity": [1, 3, 5, 1, 8, 1, 9, 3], .....: "Date": [ .....: datetime.datetime(2013, 1, 1, 13, 0), .....: datetime.datetime(2013, 1, 1, 13, 5), .....: datetime.datetime(2013, 10, 1, 20, 0), .....: datetime.datetime(2013, 10, 2, 10, 0), .....: datetime.datetime(2013, 10, 1, 20, 0), .....: datetime.datetime(2013, 10, 2, 10, 0), .....: datetime.datetime(2013, 12, 2, 12, 0), .....: datetime.datetime(2013, 12, 2, 14, 0), .....: ], .....: } .....: ) .....: In [182]: df Out[182]: Branch Buyer Quantity Date 0 A Carl 1 2013-01-01 13:00:00 1 A Mark 3 2013-01-01 13:05:00 2 A Carl 5 2013-10-01 20:00:00 3 A Carl 1 2013-10-02 10:00:00 4 A Joe 8 2013-10-01 20:00:00 5 A Joe 1 2013-10-02 10:00:00 6 A Joe 9 2013-12-02 12:00:00 7 B Carl 3 2013-12-02 14:00:00
Groupby a specific column with the desired frequency. This is like resampling.
In [183]: df.groupby([pd.Grouper(freq="1M", key="Date"), "Buyer"]).sum() Out[183]: Quantity Date Buyer 2013-01-31 Carl 1 Mark 3 2013-10-31 Carl 6 Joe 9 2013-12-31 Carl 3 Joe 9
You have an ambiguous specification in that you have a named index and a column that could be potential groupers.
In [184]: df = df.set_index("Date") In [185]: df["Date"] = df.index + pd.offsets.MonthEnd(2) In [186]: df.groupby([pd.Grouper(freq="6M", key="Date"), "Buyer"]).sum() Out[186]: Quantity Date Buyer 2013-02-28 Carl 1 Mark 3 2014-02-28 Carl 9 Joe 18 In [187]: df.groupby([pd.Grouper(freq="6M", level="Date"), "Buyer"]).sum() Out[187]: Quantity Date Buyer 2013-01-31 Carl 1 Mark 3 2014-01-31 Carl 9 Joe 18
Just like for a DataFrame or Series you can call head and tail on a groupby:
In [188]: df = pd.DataFrame([[1, 2], [1, 4], [5, 6]], columns=["A", "B"]) In [189]: df Out[189]: A B 0 1 2 1 1 4 2 5 6 In [190]: g = df.groupby("A") In [191]: g.head(1) Out[191]: A B 0 1 2 2 5 6 In [192]: g.tail(1) Out[192]: A B 1 1 4 2 5 6
This shows the first or last n rows from each group.
To select from a DataFrame or Series the nth item, use nth(). This is a reduction method, and will return a single row (or no row) per group if you pass an int for n:
In [193]: df = pd.DataFrame([[1, np.nan], [1, 4], [5, 6]], columns=["A", "B"]) In [194]: g = df.groupby("A") In [195]: g.nth(0) Out[195]: B A 1 NaN 5 6.0 In [196]: g.nth(-1) Out[196]: B A 1 4.0 5 6.0 In [197]: g.nth(1) Out[197]: B A 1 4.0
If you want to select the nth not-null item, use the dropna kwarg. For a DataFrame this should be either 'any' or 'all' just like you would pass to dropna:
'any'
'all'
# nth(0) is the same as g.first() In [198]: g.nth(0, dropna="any") Out[198]: B A 1 4.0 5 6.0 In [199]: g.first() Out[199]: B A 1 4.0 5 6.0 # nth(-1) is the same as g.last() In [200]: g.nth(-1, dropna="any") # NaNs denote group exhausted when using dropna Out[200]: B A 1 4.0 5 6.0 In [201]: g.last() Out[201]: B A 1 4.0 5 6.0 In [202]: g.B.nth(0, dropna="all") Out[202]: A 1 4.0 5 6.0 Name: B, dtype: float64
As with other methods, passing as_index=False, will achieve a filtration, which returns the grouped row.
In [203]: df = pd.DataFrame([[1, np.nan], [1, 4], [5, 6]], columns=["A", "B"]) In [204]: g = df.groupby("A", as_index=False) In [205]: g.nth(0) Out[205]: A B 0 1 NaN 2 5 6.0 In [206]: g.nth(-1) Out[206]: A B 1 1 4.0 2 5 6.0
You can also select multiple rows from each group by specifying multiple nth values as a list of ints.
In [207]: business_dates = pd.date_range(start="4/1/2014", end="6/30/2014", freq="B") In [208]: df = pd.DataFrame(1, index=business_dates, columns=["a", "b"]) # get the first, 4th, and last date index for each month In [209]: df.groupby([df.index.year, df.index.month]).nth([0, 3, -1]) Out[209]: a b 2014 4 1 1 4 1 1 4 1 1 5 1 1 5 1 1 5 1 1 6 1 1 6 1 1 6 1 1
To see the order in which each row appears within its group, use the cumcount method:
cumcount
In [210]: dfg = pd.DataFrame(list("aaabba"), columns=["A"]) In [211]: dfg Out[211]: A 0 a 1 a 2 a 3 b 4 b 5 a In [212]: dfg.groupby("A").cumcount() Out[212]: 0 0 1 1 2 2 3 0 4 1 5 3 dtype: int64 In [213]: dfg.groupby("A").cumcount(ascending=False) Out[213]: 0 3 1 2 2 1 3 1 4 0 5 0 dtype: int64
To see the ordering of the groups (as opposed to the order of rows within a group given by cumcount) you can use ngroup().
ngroup()
Note that the numbers given to the groups match the order in which the groups would be seen when iterating over the groupby object, not the order they are first observed.
In [214]: dfg = pd.DataFrame(list("aaabba"), columns=["A"]) In [215]: dfg Out[215]: A 0 a 1 a 2 a 3 b 4 b 5 a In [216]: dfg.groupby("A").ngroup() Out[216]: 0 0 1 0 2 0 3 1 4 1 5 0 dtype: int64 In [217]: dfg.groupby("A").ngroup(ascending=False) Out[217]: 0 1 1 1 2 1 3 0 4 0 5 1 dtype: int64
Groupby also works with some plotting methods. For example, suppose we suspect that some features in a DataFrame may differ by group, in this case, the values in column 1 where the group is “B” are 3 higher on average.
In [218]: np.random.seed(1234) In [219]: df = pd.DataFrame(np.random.randn(50, 2)) In [220]: df["g"] = np.random.choice(["A", "B"], size=50) In [221]: df.loc[df["g"] == "B", 1] += 3
We can easily visualize this with a boxplot:
In [222]: df.groupby("g").boxplot() Out[222]: A AxesSubplot(0.1,0.15;0.363636x0.75) B AxesSubplot(0.536364,0.15;0.363636x0.75) dtype: object
The result of calling boxplot is a dictionary whose keys are the values of our grouping column g (“A” and “B”). The values of the resulting dictionary can be controlled by the return_type keyword of boxplot. See the visualization documentation for more.
boxplot
g
return_type
For historical reasons, df.groupby("g").boxplot() is not equivalent to df.boxplot(by="g"). See here for an explanation.
df.groupby("g").boxplot()
df.boxplot(by="g")
Similar to the functionality provided by DataFrame and Series, functions that take GroupBy objects can be chained together using a pipe method to allow for a cleaner, more readable syntax. To read about .pipe in general terms, see here.
pipe
.pipe
Combining .groupby and .pipe is often useful when you need to reuse GroupBy objects.
.groupby
As an example, imagine having a DataFrame with columns for stores, products, revenue and quantity sold. We’d like to do a groupwise calculation of prices (i.e. revenue/quantity) per store and per product. We could do this in a multi-step operation, but expressing it in terms of piping can make the code more readable. First we set the data:
In [223]: n = 1000 In [224]: df = pd.DataFrame( .....: { .....: "Store": np.random.choice(["Store_1", "Store_2"], n), .....: "Product": np.random.choice(["Product_1", "Product_2"], n), .....: "Revenue": (np.random.random(n) * 50 + 10).round(2), .....: "Quantity": np.random.randint(1, 10, size=n), .....: } .....: ) .....: In [225]: df.head(2) Out[225]: Store Product Revenue Quantity 0 Store_2 Product_1 26.12 1 1 Store_2 Product_1 28.86 1
Now, to find prices per store/product, we can simply do:
In [226]: ( .....: df.groupby(["Store", "Product"]) .....: .pipe(lambda grp: grp.Revenue.sum() / grp.Quantity.sum()) .....: .unstack() .....: .round(2) .....: ) .....: Out[226]: Product Product_1 Product_2 Store Store_1 6.82 7.05 Store_2 6.30 6.64
Piping can also be expressive when you want to deliver a grouped object to some arbitrary function, for example:
In [227]: def mean(groupby): .....: return groupby.mean() .....: In [228]: df.groupby(["Store", "Product"]).pipe(mean) Out[228]: Revenue Quantity Store Product Store_1 Product_1 34.622727 5.075758 Product_2 35.482815 5.029630 Store_2 Product_1 32.972837 5.237589 Product_2 34.684360 5.224000
where mean takes a GroupBy object and finds the mean of the Revenue and Quantity columns respectively for each Store-Product combination. The mean function can be any function that takes in a GroupBy object; the .pipe will pass the GroupBy object as a parameter into the function you specify.
Regroup columns of a DataFrame according to their sum, and sum the aggregated ones.
In [229]: df = pd.DataFrame({"a": [1, 0, 0], "b": [0, 1, 0], "c": [1, 0, 0], "d": [2, 3, 4]}) In [230]: df Out[230]: a b c d 0 1 0 1 2 1 0 1 0 3 2 0 0 0 4 In [231]: df.groupby(df.sum(), axis=1).sum() Out[231]: 1 9 0 2 2 1 1 3 2 0 4
By using ngroup(), we can extract information about the groups in a way similar to factorize() (as described further in the reshaping API) but which applies naturally to multiple columns of mixed type and different sources. This can be useful as an intermediate categorical-like step in processing, when the relationships between the group rows are more important than their content, or as input to an algorithm which only accepts the integer encoding. (For more information about support in pandas for full categorical data, see the Categorical introduction and the API documentation.)
factorize()
In [232]: dfg = pd.DataFrame({"A": [1, 1, 2, 3, 2], "B": list("aaaba")}) In [233]: dfg Out[233]: A B 0 1 a 1 1 a 2 2 a 3 3 b 4 2 a In [234]: dfg.groupby(["A", "B"]).ngroup() Out[234]: 0 0 1 0 2 1 3 2 4 1 dtype: int64 In [235]: dfg.groupby(["A", [0, 0, 0, 1, 1]]).ngroup() Out[235]: 0 0 1 0 2 1 3 3 4 2 dtype: int64
Resampling produces new hypothetical samples (resamples) from already existing observed data or from a model that generates data. These new samples are similar to the pre-existing samples.
In order to resample to work on indices that are non-datetimelike, the following procedure can be utilized.
In the following examples, df.index // 5 returns a binary array which is used to determine what gets selected for the groupby operation.
The below example shows how we can downsample by consolidation of samples into fewer samples. Here by using df.index // 5, we are aggregating the samples in bins. By applying std() function, we aggregate the information contained in many samples into a small subset of values which is their standard deviation thereby reducing the number of samples.
In [236]: df = pd.DataFrame(np.random.randn(10, 2)) In [237]: df Out[237]: 0 1 0 -0.793893 0.321153 1 0.342250 1.618906 2 -0.975807 1.918201 3 -0.810847 -1.405919 4 -1.977759 0.461659 5 0.730057 -1.316938 6 -0.751328 0.528290 7 -0.257759 -1.081009 8 0.505895 -1.701948 9 -1.006349 0.020208 In [238]: df.index // 5 Out[238]: Int64Index([0, 0, 0, 0, 0, 1, 1, 1, 1, 1], dtype='int64') In [239]: df.groupby(df.index // 5).std() Out[239]: 0 1 0 0.823647 1.312912 1 0.760109 0.942941
Group DataFrame columns, compute a set of metrics and return a named Series. The Series name is used as the name for the column index. This is especially useful in conjunction with reshaping operations such as stacking in which the column index name will be used as the name of the inserted column:
In [240]: df = pd.DataFrame( .....: { .....: "a": [0, 0, 0, 0, 1, 1, 1, 1, 2, 2, 2, 2], .....: "b": [0, 0, 1, 1, 0, 0, 1, 1, 0, 0, 1, 1], .....: "c": [1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0], .....: "d": [0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 1], .....: } .....: ) .....: In [241]: def compute_metrics(x): .....: result = {"b_sum": x["b"].sum(), "c_mean": x["c"].mean()} .....: return pd.Series(result, name="metrics") .....: In [242]: result = df.groupby("a").apply(compute_metrics) In [243]: result Out[243]: metrics b_sum c_mean a 0 2.0 0.5 1 2.0 0.5 2 2.0 0.5 In [244]: result.stack() Out[244]: a metrics 0 b_sum 2.0 c_mean 0.5 1 b_sum 2.0 c_mean 0.5 2 b_sum 2.0 c_mean 0.5 dtype: float64