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 0x7fccc68bd9d0>
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 -1.882108 0.960855 -0.009459 1.016304
2013-01-02 -1.049234 0.304553 0.530838 -0.043343
2013-01-03 0.213860 -1.431571 1.205769 0.203414
2013-01-04 -0.580846 -0.552984 0.018739 -1.755353
2013-01-05 0.361730 -0.987899 0.100992 2.175861
2013-01-06 -0.498805 0.270995 -0.047999 2.116362
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 -1.882108 0.960855 -0.009459 1.016304
2013-01-02 -1.049234 0.304553 0.530838 -0.043343
2013-01-03 0.213860 -1.431571 1.205769 0.203414
2013-01-04 -0.580846 -0.552984 0.018739 -1.755353
2013-01-05 0.361730 -0.987899 0.100992 2.175861
In [15]: df.tail(3).execute()
Out[15]:
A B C D
2013-01-04 -0.580846 -0.552984 0.018739 -1.755353
2013-01-05 0.361730 -0.987899 0.100992 2.175861
2013-01-06 -0.498805 0.270995 -0.047999 2.116362
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([[-1.8821078 , 0.96085483, -0.00945897, 1.01630381],
[-1.04923352, 0.30455254, 0.53083836, -0.04334307],
[ 0.21385999, -1.43157104, 1.20576944, 0.20341422],
[-0.58084625, -0.55298387, 0.01873937, -1.75535283],
[ 0.36172962, -0.9878991 , 0.10099177, 2.1758611 ],
[-0.49880457, 0.27099466, -0.04799914, 2.11636246]])
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 -0.572567 -0.239342 0.299813 0.618874
std 0.829336 0.902979 0.491789 1.488135
min -1.882108 -1.431571 -0.047999 -1.755353
25% -0.932137 -0.879170 -0.002409 0.018346
50% -0.539825 -0.140995 0.059866 0.609859
75% 0.035694 0.296163 0.423377 1.841348
max 0.361730 0.960855 1.205769 2.175861
Sorting by an axis:
In [21]: df.sort_index(axis=1, ascending=False).execute()
Out[21]:
D C B A
2013-01-01 1.016304 -0.009459 0.960855 -1.882108
2013-01-02 -0.043343 0.530838 0.304553 -1.049234
2013-01-03 0.203414 1.205769 -1.431571 0.213860
2013-01-04 -1.755353 0.018739 -0.552984 -0.580846
2013-01-05 2.175861 0.100992 -0.987899 0.361730
2013-01-06 2.116362 -0.047999 0.270995 -0.498805
Sorting by values:
In [22]: df.sort_values(by='B').execute()
Out[22]:
A B C D
2013-01-03 0.213860 -1.431571 1.205769 0.203414
2013-01-05 0.361730 -0.987899 0.100992 2.175861
2013-01-04 -0.580846 -0.552984 0.018739 -1.755353
2013-01-06 -0.498805 0.270995 -0.047999 2.116362
2013-01-02 -1.049234 0.304553 0.530838 -0.043343
2013-01-01 -1.882108 0.960855 -0.009459 1.016304
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 -1.882108
2013-01-02 -1.049234
2013-01-03 0.213860
2013-01-04 -0.580846
2013-01-05 0.361730
2013-01-06 -0.498805
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 -1.882108 0.960855 -0.009459 1.016304
2013-01-02 -1.049234 0.304553 0.530838 -0.043343
2013-01-03 0.213860 -1.431571 1.205769 0.203414
In [25]: df['20130102':'20130104'].execute()
Out[25]:
A B C D
2013-01-02 -1.049234 0.304553 0.530838 -0.043343
2013-01-03 0.213860 -1.431571 1.205769 0.203414
2013-01-04 -0.580846 -0.552984 0.018739 -1.755353
Selection by label#
For getting a cross section using a label:
In [26]: df.loc['20130101'].execute()
Out[26]:
A -1.882108
B 0.960855
C -0.009459
D 1.016304
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 -1.882108 0.960855
2013-01-02 -1.049234 0.304553
2013-01-03 0.213860 -1.431571
2013-01-04 -0.580846 -0.552984
2013-01-05 0.361730 -0.987899
2013-01-06 -0.498805 0.270995
Showing label slicing, both endpoints are included:
In [28]: df.loc['20130102':'20130104', ['A', 'B']].execute()
Out[28]:
A B
2013-01-02 -1.049234 0.304553
2013-01-03 0.213860 -1.431571
2013-01-04 -0.580846 -0.552984
Reduction in the dimensions of the returned object:
In [29]: df.loc['20130102', ['A', 'B']].execute()
Out[29]:
A -1.049234
B 0.304553
Name: 2013-01-02 00:00:00, dtype: float64
For getting a scalar value:
In [30]: df.loc['20130101', 'A'].execute()
Out[30]: -1.8821077980771692
For getting fast access to a scalar (equivalent to the prior method):
In [31]: df.at['20130101', 'A'].execute()
Out[31]: -1.8821077980771692
Selection by position#
Select via the position of the passed integers:
In [32]: df.iloc[3].execute()
Out[32]:
A -0.580846
B -0.552984
C 0.018739
D -1.755353
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 -0.580846 -0.552984
2013-01-05 0.361730 -0.987899
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 -1.049234 0.530838
2013-01-03 0.213860 1.205769
2013-01-05 0.361730 0.100992
For slicing rows explicitly:
In [35]: df.iloc[1:3, :].execute()
Out[35]:
A B C D
2013-01-02 -1.049234 0.304553 0.530838 -0.043343
2013-01-03 0.213860 -1.431571 1.205769 0.203414
For slicing columns explicitly:
In [36]: df.iloc[:, 1:3].execute()
Out[36]:
B C
2013-01-01 0.960855 -0.009459
2013-01-02 0.304553 0.530838
2013-01-03 -1.431571 1.205769
2013-01-04 -0.552984 0.018739
2013-01-05 -0.987899 0.100992
2013-01-06 0.270995 -0.047999
For getting a value explicitly:
In [37]: df.iloc[1, 1].execute()
Out[37]: 0.30455253590692033
For getting fast access to a scalar (equivalent to the prior method):
In [38]: df.iat[1, 1].execute()
Out[38]: 0.30455253590692033
Boolean indexing#
Using a single column’s values to select data.
In [39]: df[df['A'] > 0].execute()
Out[39]:
A B C D
2013-01-03 0.21386 -1.431571 1.205769 0.203414
2013-01-05 0.36173 -0.987899 0.100992 2.175861
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 NaN 0.960855 NaN 1.016304
2013-01-02 NaN 0.304553 0.530838 NaN
2013-01-03 0.21386 NaN 1.205769 0.203414
2013-01-04 NaN NaN 0.018739 NaN
2013-01-05 0.36173 NaN 0.100992 2.175861
2013-01-06 NaN 0.270995 NaN 2.116362
Operations#
Stats#
Operations in general exclude missing data.
Performing a descriptive statistic:
In [41]: df.mean().execute()
Out[41]:
A -0.572567
B -0.239342
C 0.299813
D 0.618874
dtype: float64
Same operation on the other axis:
In [42]: df.mean(1).execute()
Out[42]:
2013-01-01 0.021398
2013-01-02 -0.064296
2013-01-03 0.047868
2013-01-04 -0.717611
2013-01-05 0.412671
2013-01-06 0.460138
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 -0.786140 -2.431571 0.205769 -0.796586
2013-01-04 -3.580846 -3.552984 -2.981261 -4.755353
2013-01-05 -4.638270 -5.987899 -4.899008 -2.824139
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 2.243837
B 2.392426
C 1.253769
D 3.931214
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 0.987942 -0.929619 -0.066542 -2.142220
1 0.612161 1.606244 -1.309438 0.076113
2 -0.419658 -0.289739 -1.657454 -0.115085
3 1.953100 0.686447 -1.513060 0.598228
4 0.134242 -0.313226 1.966364 -0.069673
5 0.456217 0.279538 0.440514 -0.220514
6 -0.479430 -0.648801 -0.841180 -0.892234
7 -0.662661 -0.059356 0.375136 0.202591
8 1.602243 0.110080 -0.699141 -1.101087
9 0.673317 0.431241 -0.048495 0.201615
# 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 0.987942 -0.929619 -0.066542 -2.142220
1 0.612161 1.606244 -1.309438 0.076113
2 -0.419658 -0.289739 -1.657454 -0.115085
3 1.953100 0.686447 -1.513060 0.598228
4 0.134242 -0.313226 1.966364 -0.069673
5 0.456217 0.279538 0.440514 -0.220514
6 -0.479430 -0.648801 -0.841180 -0.892234
7 -0.662661 -0.059356 0.375136 0.202591
8 1.602243 0.110080 -0.699141 -1.101087
9 0.673317 0.431241 -0.048495 0.201615
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 1.299919 0.249839
1 bar one 0.641778 0.430571
2 foo two 2.248535 -2.493165
3 bar three 0.486686 -0.901611
4 foo two 0.614228 0.297693
5 bar two 0.993868 0.086885
6 foo one -0.271946 -0.922745
7 foo three -1.713409 1.831023
Grouping and then applying the sum()
function to the resulting
groups.
In [65]: df.groupby('A').sum().execute()
Out[65]:
C D
A
bar 2.122332 -0.384155
foo 2.177327 -1.037355
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 0.641778 0.430571
three 0.486686 -0.901611
two 0.993868 0.086885
foo one 1.027973 -0.672906
three -1.713409 1.831023
two 2.862763 -2.195472
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 0x7fccc6a99650>

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 -0.657630 -1.628396 -0.521589 -0.628873
1 2000-01-02 -0.618936 -2.408207 0.374808 -3.626958
2 2000-01-03 0.738739 -2.872245 1.326673 -3.479289
3 2000-01-04 1.906156 -2.158316 0.062554 -3.765099
4 2000-01-05 2.149986 -2.311702 0.637701 -3.184967
.. ... ... ... ... ...
995 2002-09-22 21.214772 -42.538924 -42.710949 32.937965
996 2002-09-23 20.857961 -42.768397 -42.648203 31.736822
997 2002-09-24 21.377999 -42.700526 -41.447758 32.803241
998 2002-09-25 22.125392 -43.409634 -41.340038 32.609549
999 2002-09-26 21.517450 -43.961246 -40.317388 32.710258
[1000 rows x 5 columns]