10 minutes to Mars DataFrame#
This is a short introduction to Mars DataFrame which is originated from 10 minutes to pandas.
Customarily, we import as follows:
In [1]: import mars
In [2]: import mars.tensor as mt
In [3]: import mars.dataframe as md
Now create a new default session.
In [4]: mars.new_session()
Out[4]: <mars.deploy.oscar.session.SyncSession at 0x7f520fa9bf50>
Object creation#
Creating a Series
by passing a list of values, letting it create
a default integer index:
In [5]: s = md.Series([1, 3, 5, mt.nan, 6, 8])
In [6]: s.execute()
Out[6]:
0 1.0
1 3.0
2 5.0
3 NaN
4 6.0
5 8.0
dtype: float64
Creating a DataFrame
by passing a Mars tensor, with a datetime index
and labeled columns:
In [7]: dates = md.date_range('20130101', periods=6)
In [8]: dates.execute()
Out[8]:
DatetimeIndex(['2013-01-01', '2013-01-02', '2013-01-03', '2013-01-04',
'2013-01-05', '2013-01-06'],
dtype='datetime64[ns]', freq='D')
In [9]: df = md.DataFrame(mt.random.randn(6, 4), index=dates, columns=list('ABCD'))
In [10]: df.execute()
Out[10]:
A B C D
2013-01-01 0.131202 0.806436 -1.017856 -1.098031
2013-01-02 1.353260 -0.751281 -0.676277 -0.097598
2013-01-03 -1.340585 -0.823038 1.587496 -1.234449
2013-01-04 0.265652 -0.002158 2.200595 2.004409
2013-01-05 0.169342 -0.251483 0.153520 -0.362017
2013-01-06 0.197870 0.107158 -0.139387 0.652527
Creating a DataFrame
by passing a dict of objects that can be converted to series-like.
In [11]: df2 = md.DataFrame({'A': 1.,
....: 'B': md.Timestamp('20130102'),
....: 'C': md.Series(1, index=list(range(4)), dtype='float32'),
....: 'D': mt.array([3] * 4, dtype='int32'),
....: 'E': 'foo'})
....:
In [12]: df2.execute()
Out[12]:
A B C D E
0 1.0 2013-01-02 1.0 3 foo
1 1.0 2013-01-02 1.0 3 foo
2 1.0 2013-01-02 1.0 3 foo
3 1.0 2013-01-02 1.0 3 foo
The columns of the resulting DataFrame
have different dtypes.
In [13]: df2.dtypes
Out[13]:
A float64
B datetime64[ns]
C float32
D int32
E object
dtype: object
Viewing data#
Here is how to view the top and bottom rows of the frame:
In [14]: df.head().execute()
Out[14]:
A B C D
2013-01-01 0.131202 0.806436 -1.017856 -1.098031
2013-01-02 1.353260 -0.751281 -0.676277 -0.097598
2013-01-03 -1.340585 -0.823038 1.587496 -1.234449
2013-01-04 0.265652 -0.002158 2.200595 2.004409
2013-01-05 0.169342 -0.251483 0.153520 -0.362017
In [15]: df.tail(3).execute()
Out[15]:
A B C D
2013-01-04 0.265652 -0.002158 2.200595 2.004409
2013-01-05 0.169342 -0.251483 0.153520 -0.362017
2013-01-06 0.197870 0.107158 -0.139387 0.652527
Display the index, columns:
In [16]: df.index.execute()
Out[16]:
DatetimeIndex(['2013-01-01', '2013-01-02', '2013-01-03', '2013-01-04',
'2013-01-05', '2013-01-06'],
dtype='datetime64[ns]', freq='D')
In [17]: df.columns.execute()
Out[17]: Index(['A', 'B', 'C', 'D'], dtype='object')
DataFrame.to_tensor()
gives a Mars tensor representation of the underlying data.
Note that this can be an expensive operation when your DataFrame
has
columns with different data types, which comes down to a fundamental difference
between DataFrame and tensor: tensors have one dtype for the entire tensor,
while DataFrames have one dtype per column. When you call
DataFrame.to_tensor()
, Mars DataFrame will find the tensor dtype that can hold all
of the dtypes in the DataFrame. This may end up being object
, which requires
casting every value to a Python object.
For df
, our DataFrame
of all floating-point values,
DataFrame.to_tensor()
is fast and doesn’t require copying data.
In [18]: df.to_tensor().execute()
Out[18]:
array([[ 1.31201729e-01, 8.06435653e-01, -1.01785596e+00,
-1.09803098e+00],
[ 1.35326041e+00, -7.51280648e-01, -6.76276948e-01,
-9.75979588e-02],
[-1.34058477e+00, -8.23037990e-01, 1.58749612e+00,
-1.23444884e+00],
[ 2.65652215e-01, -2.15804872e-03, 2.20059480e+00,
2.00440905e+00],
[ 1.69341795e-01, -2.51482962e-01, 1.53520168e-01,
-3.62017238e-01],
[ 1.97869804e-01, 1.07158085e-01, -1.39387347e-01,
6.52526637e-01]])
For df2
, the DataFrame
with multiple dtypes,
DataFrame.to_tensor()
is relatively expensive.
In [19]: df2.to_tensor().execute()
Out[19]:
array([[1.0, Timestamp('2013-01-02 00:00:00'), 1.0, 3, 'foo'],
[1.0, Timestamp('2013-01-02 00:00:00'), 1.0, 3, 'foo'],
[1.0, Timestamp('2013-01-02 00:00:00'), 1.0, 3, 'foo'],
[1.0, Timestamp('2013-01-02 00:00:00'), 1.0, 3, 'foo']],
dtype=object)
Note
DataFrame.to_tensor()
does not include the index or column
labels in the output.
describe()
shows a quick statistic summary of your data:
In [20]: df.describe().execute()
Out[20]:
A B C D
count 6.000000 6.000000 6.000000 6.000000
mean 0.129457 -0.152394 0.351348 -0.022527
std 0.858317 0.604573 1.277377 1.209175
min -1.340585 -0.823038 -1.017856 -1.234449
25% 0.140737 -0.626331 -0.542055 -0.914028
50% 0.183606 -0.126821 0.007066 -0.229808
75% 0.248707 0.079829 1.229002 0.464995
max 1.353260 0.806436 2.200595 2.004409
Sorting by an axis:
In [21]: df.sort_index(axis=1, ascending=False).execute()
Out[21]:
D C B A
2013-01-01 -1.098031 -1.017856 0.806436 0.131202
2013-01-02 -0.097598 -0.676277 -0.751281 1.353260
2013-01-03 -1.234449 1.587496 -0.823038 -1.340585
2013-01-04 2.004409 2.200595 -0.002158 0.265652
2013-01-05 -0.362017 0.153520 -0.251483 0.169342
2013-01-06 0.652527 -0.139387 0.107158 0.197870
Sorting by values:
In [22]: df.sort_values(by='B').execute()
Out[22]:
A B C D
2013-01-03 -1.340585 -0.823038 1.587496 -1.234449
2013-01-02 1.353260 -0.751281 -0.676277 -0.097598
2013-01-05 0.169342 -0.251483 0.153520 -0.362017
2013-01-04 0.265652 -0.002158 2.200595 2.004409
2013-01-06 0.197870 0.107158 -0.139387 0.652527
2013-01-01 0.131202 0.806436 -1.017856 -1.098031
Selection#
Note
While standard Python / Numpy expressions for selecting and setting are
intuitive and come in handy for interactive work, for production code, we
recommend the optimized DataFrame data access methods, .at
, .iat
,
.loc
and .iloc
.
Getting#
Selecting a single column, which yields a Series
,
equivalent to df.A
:
In [23]: df['A'].execute()
Out[23]:
2013-01-01 0.131202
2013-01-02 1.353260
2013-01-03 -1.340585
2013-01-04 0.265652
2013-01-05 0.169342
2013-01-06 0.197870
Freq: D, Name: A, dtype: float64
Selecting via []
, which slices the rows.
In [24]: df[0:3].execute()
Out[24]:
A B C D
2013-01-01 0.131202 0.806436 -1.017856 -1.098031
2013-01-02 1.353260 -0.751281 -0.676277 -0.097598
2013-01-03 -1.340585 -0.823038 1.587496 -1.234449
In [25]: df['20130102':'20130104'].execute()
Out[25]:
A B C D
2013-01-02 1.353260 -0.751281 -0.676277 -0.097598
2013-01-03 -1.340585 -0.823038 1.587496 -1.234449
2013-01-04 0.265652 -0.002158 2.200595 2.004409
Selection by label#
For getting a cross section using a label:
In [26]: df.loc['20130101'].execute()
Out[26]:
A 0.131202
B 0.806436
C -1.017856
D -1.098031
Name: 2013-01-01 00:00:00, dtype: float64
Selecting on a multi-axis by label:
In [27]: df.loc[:, ['A', 'B']].execute()
Out[27]:
A B
2013-01-01 0.131202 0.806436
2013-01-02 1.353260 -0.751281
2013-01-03 -1.340585 -0.823038
2013-01-04 0.265652 -0.002158
2013-01-05 0.169342 -0.251483
2013-01-06 0.197870 0.107158
Showing label slicing, both endpoints are included:
In [28]: df.loc['20130102':'20130104', ['A', 'B']].execute()
Out[28]:
A B
2013-01-02 1.353260 -0.751281
2013-01-03 -1.340585 -0.823038
2013-01-04 0.265652 -0.002158
Reduction in the dimensions of the returned object:
In [29]: df.loc['20130102', ['A', 'B']].execute()
Out[29]:
A 1.353260
B -0.751281
Name: 2013-01-02 00:00:00, dtype: float64
For getting a scalar value:
In [30]: df.loc['20130101', 'A'].execute()
Out[30]: 0.13120172881188916
For getting fast access to a scalar (equivalent to the prior method):
In [31]: df.at['20130101', 'A'].execute()
Out[31]: 0.13120172881188916
Selection by position#
Select via the position of the passed integers:
In [32]: df.iloc[3].execute()
Out[32]:
A 0.265652
B -0.002158
C 2.200595
D 2.004409
Name: 2013-01-04 00:00:00, dtype: float64
By integer slices, acting similar to numpy/python:
In [33]: df.iloc[3:5, 0:2].execute()
Out[33]:
A B
2013-01-04 0.265652 -0.002158
2013-01-05 0.169342 -0.251483
By lists of integer position locations, similar to the numpy/python style:
In [34]: df.iloc[[1, 2, 4], [0, 2]].execute()
Out[34]:
A C
2013-01-02 1.353260 -0.676277
2013-01-03 -1.340585 1.587496
2013-01-05 0.169342 0.153520
For slicing rows explicitly:
In [35]: df.iloc[1:3, :].execute()
Out[35]:
A B C D
2013-01-02 1.353260 -0.751281 -0.676277 -0.097598
2013-01-03 -1.340585 -0.823038 1.587496 -1.234449
For slicing columns explicitly:
In [36]: df.iloc[:, 1:3].execute()
Out[36]:
B C
2013-01-01 0.806436 -1.017856
2013-01-02 -0.751281 -0.676277
2013-01-03 -0.823038 1.587496
2013-01-04 -0.002158 2.200595
2013-01-05 -0.251483 0.153520
2013-01-06 0.107158 -0.139387
For getting a value explicitly:
In [37]: df.iloc[1, 1].execute()
Out[37]: -0.7512806482972055
For getting fast access to a scalar (equivalent to the prior method):
In [38]: df.iat[1, 1].execute()
Out[38]: -0.7512806482972055
Boolean indexing#
Using a single column’s values to select data.
In [39]: df[df['A'] > 0].execute()
Out[39]:
A B C D
2013-01-01 0.131202 0.806436 -1.017856 -1.098031
2013-01-02 1.353260 -0.751281 -0.676277 -0.097598
2013-01-04 0.265652 -0.002158 2.200595 2.004409
2013-01-05 0.169342 -0.251483 0.153520 -0.362017
2013-01-06 0.197870 0.107158 -0.139387 0.652527
Selecting values from a DataFrame where a boolean condition is met.
In [40]: df[df > 0].execute()
Out[40]:
A B C D
2013-01-01 0.131202 0.806436 NaN NaN
2013-01-02 1.353260 NaN NaN NaN
2013-01-03 NaN NaN 1.587496 NaN
2013-01-04 0.265652 NaN 2.200595 2.004409
2013-01-05 0.169342 NaN 0.153520 NaN
2013-01-06 0.197870 0.107158 NaN 0.652527
Operations#
Stats#
Operations in general exclude missing data.
Performing a descriptive statistic:
In [41]: df.mean().execute()
Out[41]:
A 0.129457
B -0.152394
C 0.351348
D -0.022527
dtype: float64
Same operation on the other axis:
In [42]: df.mean(1).execute()
Out[42]:
2013-01-01 -0.294562
2013-01-02 -0.042974
2013-01-03 -0.452644
2013-01-04 1.117125
2013-01-05 -0.072660
2013-01-06 0.204542
Freq: D, dtype: float64
Operating with objects that have different dimensionality and need alignment. In addition, Mars DataFrame automatically broadcasts along the specified dimension.
In [43]: s = md.Series([1, 3, 5, mt.nan, 6, 8], index=dates).shift(2)
In [44]: s.execute()
Out[44]:
2013-01-01 NaN
2013-01-02 NaN
2013-01-03 1.0
2013-01-04 3.0
2013-01-05 5.0
2013-01-06 NaN
Freq: D, dtype: float64
In [45]: df.sub(s, axis='index').execute()
Out[45]:
A B C D
2013-01-01 NaN NaN NaN NaN
2013-01-02 NaN NaN NaN NaN
2013-01-03 -2.340585 -1.823038 0.587496 -2.234449
2013-01-04 -2.734348 -3.002158 -0.799405 -0.995591
2013-01-05 -4.830658 -5.251483 -4.846480 -5.362017
2013-01-06 NaN NaN NaN NaN
Apply#
Applying functions to the data:
In [46]: df.apply(lambda x: x.max() - x.min()).execute()
Out[46]:
A 2.693845
B 1.629474
C 3.218451
D 3.238858
dtype: float64
String Methods#
Series is equipped with a set of string processing methods in the str attribute that make it easy to operate on each element of the array, as in the code snippet below. Note that pattern-matching in str generally uses regular expressions by default (and in some cases always uses them). See more at Vectorized String Methods.
In [47]: s = md.Series(['A', 'B', 'C', 'Aaba', 'Baca', mt.nan, 'CABA', 'dog', 'cat'])
In [48]: s.str.lower().execute()
Out[48]:
0 a
1 b
2 c
3 aaba
4 baca
5 NaN
6 caba
7 dog
8 cat
dtype: object
Merge#
Concat#
Mars DataFrame provides various facilities for easily combining together Series and DataFrame objects with various kinds of set logic for the indexes and relational algebra functionality in the case of join / merge-type operations.
Concatenating DataFrame objects together with concat()
:
In [49]: df = md.DataFrame(mt.random.randn(10, 4))
In [50]: df.execute()
Out[50]:
0 1 2 3
0 0.340665 -0.017870 2.639970 -0.661813
1 1.658088 1.856396 -1.750673 -0.167159
2 1.110126 0.196777 -0.352805 -0.196033
3 0.084739 0.345062 0.198985 -0.202668
4 0.054834 -0.603046 1.132117 -0.184715
5 1.224022 1.182662 0.047055 -0.422539
6 -0.625837 0.841115 -0.295166 -0.800794
7 -1.819517 1.898076 0.933760 -0.396641
8 -0.982989 -1.757011 -0.515680 -0.810376
9 -0.917127 -0.092858 -0.501922 -0.471996
# break it into pieces
In [51]: pieces = [df[:3], df[3:7], df[7:]]
In [52]: md.concat(pieces).execute()
Out[52]:
0 1 2 3
0 0.340665 -0.017870 2.639970 -0.661813
1 1.658088 1.856396 -1.750673 -0.167159
2 1.110126 0.196777 -0.352805 -0.196033
3 0.084739 0.345062 0.198985 -0.202668
4 0.054834 -0.603046 1.132117 -0.184715
5 1.224022 1.182662 0.047055 -0.422539
6 -0.625837 0.841115 -0.295166 -0.800794
7 -1.819517 1.898076 0.933760 -0.396641
8 -0.982989 -1.757011 -0.515680 -0.810376
9 -0.917127 -0.092858 -0.501922 -0.471996
Join#
SQL style merges. See the Database style joining section.
In [53]: left = md.DataFrame({'key': ['foo', 'foo'], 'lval': [1, 2]})
In [54]: right = md.DataFrame({'key': ['foo', 'foo'], 'rval': [4, 5]})
In [55]: left.execute()
Out[55]:
key lval
0 foo 1
1 foo 2
In [56]: right.execute()
Out[56]:
key rval
0 foo 4
1 foo 5
In [57]: md.merge(left, right, on='key').execute()
Out[57]:
key lval rval
0 foo 1 4
1 foo 1 5
2 foo 2 4
3 foo 2 5
Another example that can be given is:
In [58]: left = md.DataFrame({'key': ['foo', 'bar'], 'lval': [1, 2]})
In [59]: right = md.DataFrame({'key': ['foo', 'bar'], 'rval': [4, 5]})
In [60]: left.execute()
Out[60]:
key lval
0 foo 1
1 bar 2
In [61]: right.execute()
Out[61]:
key rval
0 foo 4
1 bar 5
In [62]: md.merge(left, right, on='key').execute()
Out[62]:
key lval rval
0 foo 1 4
1 bar 2 5
Grouping#
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
In [63]: df = md.DataFrame({'A': ['foo', 'bar', 'foo', 'bar',
....: 'foo', 'bar', 'foo', 'foo'],
....: 'B': ['one', 'one', 'two', 'three',
....: 'two', 'two', 'one', 'three'],
....: 'C': mt.random.randn(8),
....: 'D': mt.random.randn(8)})
....:
In [64]: df.execute()
Out[64]:
A B C D
0 foo one 0.202711 -0.448300
1 bar one -0.985369 -1.963379
2 foo two -0.061727 0.417080
3 bar three -1.942882 0.664544
4 foo two -0.138829 -1.339837
5 bar two 0.159911 -0.667307
6 foo one -1.225715 -1.105817
7 foo three -1.378385 1.021276
Grouping and then applying the sum()
function to the resulting
groups.
In [65]: df.groupby('A').sum().execute()
Out[65]:
C D
A
bar -2.768340 -1.966141
foo -2.601944 -1.455598
Grouping by multiple columns forms a hierarchical index, and again we can apply the sum function.
In [66]: df.groupby(['A', 'B']).sum().execute()
Out[66]:
C D
A B
bar one -0.985369 -1.963379
three -1.942882 0.664544
two 0.159911 -0.667307
foo one -1.023003 -1.554117
three -1.378385 1.021276
two -0.200556 -0.922757
Plotting#
We use the standard convention for referencing the matplotlib API:
In [67]: import matplotlib.pyplot as plt
In [68]: plt.close('all')
In [69]: ts = md.Series(mt.random.randn(1000),
....: index=md.date_range('1/1/2000', periods=1000))
....:
In [70]: ts = ts.cumsum()
In [71]: ts.plot()
Out[71]: <AxesSubplot:>
On a DataFrame, the plot()
method is a convenience to plot all
of the columns with labels:
In [72]: df = md.DataFrame(mt.random.randn(1000, 4), index=ts.index,
....: columns=['A', 'B', 'C', 'D'])
....:
In [73]: df = df.cumsum()
In [74]: plt.figure()
Out[74]: <Figure size 640x480 with 0 Axes>
In [75]: df.plot()
Out[75]: <AxesSubplot:>
In [76]: plt.legend(loc='best')
Out[76]: <matplotlib.legend.Legend at 0x7f52185464d0>
Getting data in/out#
CSV#
In [77]: df.to_csv('foo.csv').execute()
Out[77]:
Empty DataFrame
Columns: []
Index: []
In [78]: md.read_csv('foo.csv').execute()
Out[78]:
Unnamed: 0 A B C D
0 2000-01-01 0.939091 -1.039370 1.186373 -0.435623
1 2000-01-02 1.015892 -1.445657 0.775589 -0.708444
2 2000-01-03 0.126036 0.038574 0.772313 -1.454835
3 2000-01-04 -1.923846 -0.701713 2.196586 -0.945821
4 2000-01-05 -0.561005 1.171094 0.708817 0.787630
.. ... ... ... ... ...
995 2002-09-22 -1.575408 15.838433 -24.230602 65.672469
996 2002-09-23 -1.636742 18.760213 -25.933767 66.910663
997 2002-09-24 -2.154960 19.675557 -25.497723 66.363891
998 2002-09-25 -4.315121 20.019056 -25.150312 66.148998
999 2002-09-26 -4.281877 20.081122 -24.974266 66.418760
[1000 rows x 5 columns]