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 0x7f26ddefcad0>
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 0.612509 -0.173117 2.714119 1.994359
2013-01-02 0.667038 0.320303 -0.848741 0.292421
2013-01-03 0.564992 1.278180 1.974249 -1.525575
2013-01-04 0.158850 -0.672767 -0.368342 0.258201
2013-01-05 1.399574 0.968248 0.917410 -0.265619
2013-01-06 -1.357295 -0.410361 0.313132 0.698699
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 0.612509 -0.173117 2.714119 1.994359
2013-01-02 0.667038 0.320303 -0.848741 0.292421
2013-01-03 0.564992 1.278180 1.974249 -1.525575
2013-01-04 0.158850 -0.672767 -0.368342 0.258201
2013-01-05 1.399574 0.968248 0.917410 -0.265619
In [15]: df.tail(3).execute()
Out[15]:
A B C D
2013-01-04 0.158850 -0.672767 -0.368342 0.258201
2013-01-05 1.399574 0.968248 0.917410 -0.265619
2013-01-06 -1.357295 -0.410361 0.313132 0.698699
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([[ 0.61250867, -0.17311674, 2.71411906, 1.99435889],
[ 0.66703798, 0.32030308, -0.84874085, 0.29242138],
[ 0.56499169, 1.2781802 , 1.97424878, -1.52557508],
[ 0.15884989, -0.67276679, -0.36834228, 0.25820094],
[ 1.39957398, 0.9682483 , 0.91741035, -0.2656194 ],
[-1.35729515, -0.41036109, 0.31313174, 0.69869872]])
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 0.340945 0.218414 0.783638 0.242081
std 0.923921 0.779563 1.369328 1.154482
min -1.357295 -0.672767 -0.848741 -1.525575
25% 0.260385 -0.351050 -0.197974 -0.134664
50% 0.588750 0.073593 0.615271 0.275311
75% 0.653406 0.806262 1.710039 0.597129
max 1.399574 1.278180 2.714119 1.994359
Sorting by an axis:
In [21]: df.sort_index(axis=1, ascending=False).execute()
Out[21]:
D C B A
2013-01-01 1.994359 2.714119 -0.173117 0.612509
2013-01-02 0.292421 -0.848741 0.320303 0.667038
2013-01-03 -1.525575 1.974249 1.278180 0.564992
2013-01-04 0.258201 -0.368342 -0.672767 0.158850
2013-01-05 -0.265619 0.917410 0.968248 1.399574
2013-01-06 0.698699 0.313132 -0.410361 -1.357295
Sorting by values:
In [22]: df.sort_values(by='B').execute()
Out[22]:
A B C D
2013-01-04 0.158850 -0.672767 -0.368342 0.258201
2013-01-06 -1.357295 -0.410361 0.313132 0.698699
2013-01-01 0.612509 -0.173117 2.714119 1.994359
2013-01-02 0.667038 0.320303 -0.848741 0.292421
2013-01-05 1.399574 0.968248 0.917410 -0.265619
2013-01-03 0.564992 1.278180 1.974249 -1.525575
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 0.612509
2013-01-02 0.667038
2013-01-03 0.564992
2013-01-04 0.158850
2013-01-05 1.399574
2013-01-06 -1.357295
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 0.612509 -0.173117 2.714119 1.994359
2013-01-02 0.667038 0.320303 -0.848741 0.292421
2013-01-03 0.564992 1.278180 1.974249 -1.525575
In [25]: df['20130102':'20130104'].execute()
Out[25]:
A B C D
2013-01-02 0.667038 0.320303 -0.848741 0.292421
2013-01-03 0.564992 1.278180 1.974249 -1.525575
2013-01-04 0.158850 -0.672767 -0.368342 0.258201
Selection by label¶
For getting a cross section using a label:
In [26]: df.loc['20130101'].execute()
Out[26]:
A 0.612509
B -0.173117
C 2.714119
D 1.994359
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 0.612509 -0.173117
2013-01-02 0.667038 0.320303
2013-01-03 0.564992 1.278180
2013-01-04 0.158850 -0.672767
2013-01-05 1.399574 0.968248
2013-01-06 -1.357295 -0.410361
Showing label slicing, both endpoints are included:
In [28]: df.loc['20130102':'20130104', ['A', 'B']].execute()
Out[28]:
A B
2013-01-02 0.667038 0.320303
2013-01-03 0.564992 1.278180
2013-01-04 0.158850 -0.672767
Reduction in the dimensions of the returned object:
In [29]: df.loc['20130102', ['A', 'B']].execute()
Out[29]:
A 0.667038
B 0.320303
Name: 2013-01-02 00:00:00, dtype: float64
For getting a scalar value:
In [30]: df.loc['20130101', 'A'].execute()
Out[30]: 0.6125086694517746
For getting fast access to a scalar (equivalent to the prior method):
In [31]: df.at['20130101', 'A'].execute()
Out[31]: 0.6125086694517746
Selection by position¶
Select via the position of the passed integers:
In [32]: df.iloc[3].execute()
Out[32]:
A 0.158850
B -0.672767
C -0.368342
D 0.258201
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 0.158850 -0.672767
2013-01-05 1.399574 0.968248
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 0.667038 -0.848741
2013-01-03 0.564992 1.974249
2013-01-05 1.399574 0.917410
For slicing rows explicitly:
In [35]: df.iloc[1:3, :].execute()
Out[35]:
A B C D
2013-01-02 0.667038 0.320303 -0.848741 0.292421
2013-01-03 0.564992 1.278180 1.974249 -1.525575
For slicing columns explicitly:
In [36]: df.iloc[:, 1:3].execute()
Out[36]:
B C
2013-01-01 -0.173117 2.714119
2013-01-02 0.320303 -0.848741
2013-01-03 1.278180 1.974249
2013-01-04 -0.672767 -0.368342
2013-01-05 0.968248 0.917410
2013-01-06 -0.410361 0.313132
For getting a value explicitly:
In [37]: df.iloc[1, 1].execute()
Out[37]: 0.32030308078473263
For getting fast access to a scalar (equivalent to the prior method):
In [38]: df.iat[1, 1].execute()
Out[38]: 0.32030308078473263
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 0.612509 -0.173117 2.714119 1.994359
2013-01-02 0.667038 0.320303 -0.848741 0.292421
2013-01-03 0.564992 1.278180 1.974249 -1.525575
2013-01-04 0.158850 -0.672767 -0.368342 0.258201
2013-01-05 1.399574 0.968248 0.917410 -0.265619
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 0.612509 NaN 2.714119 1.994359
2013-01-02 0.667038 0.320303 NaN 0.292421
2013-01-03 0.564992 1.278180 1.974249 NaN
2013-01-04 0.158850 NaN NaN 0.258201
2013-01-05 1.399574 0.968248 0.917410 NaN
2013-01-06 NaN NaN 0.313132 0.698699
Operations¶
Stats¶
Operations in general exclude missing data.
Performing a descriptive statistic:
In [41]: df.mean().execute()
Out[41]:
A 0.340945
B 0.218414
C 0.783638
D 0.242081
dtype: float64
Same operation on the other axis:
In [42]: df.mean(1).execute()
Out[42]:
2013-01-01 1.286967
2013-01-02 0.107755
2013-01-03 0.572961
2013-01-04 -0.156015
2013-01-05 0.754903
2013-01-06 -0.188956
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 -0.435008 0.278180 0.974249 -2.525575
2013-01-04 -2.841150 -3.672767 -3.368342 -2.741799
2013-01-05 -3.600426 -4.031752 -4.082590 -5.265619
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 2.756869
B 1.950947
C 3.562860
D 3.519934
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 0.392163 -1.180503 0.308954 -0.476383
1 -0.643694 0.330282 0.233983 -1.858699
2 -0.980115 -2.164547 -1.450455 0.486994
3 -0.311438 0.530217 -0.547803 -0.695146
4 0.320900 0.383052 0.338320 0.221177
5 -1.321418 -0.901024 0.931401 0.768191
6 0.219049 0.809248 -1.206234 1.664526
7 0.580689 -0.844961 -1.636815 -0.677618
8 0.335160 0.717138 -1.286026 1.385991
9 1.070349 -2.910198 -1.412053 -1.318169
# 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 0.392163 -1.180503 0.308954 -0.476383
1 -0.643694 0.330282 0.233983 -1.858699
2 -0.980115 -2.164547 -1.450455 0.486994
3 -0.311438 0.530217 -0.547803 -0.695146
4 0.320900 0.383052 0.338320 0.221177
5 -1.321418 -0.901024 0.931401 0.768191
6 0.219049 0.809248 -1.206234 1.664526
7 0.580689 -0.844961 -1.636815 -0.677618
8 0.335160 0.717138 -1.286026 1.385991
9 1.070349 -2.910198 -1.412053 -1.318169
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 0.836846 0.233225
1 bar one -2.435075 -1.399957
2 foo two -1.042548 1.805187
3 bar three 1.350988 0.327271
4 foo two 0.873994 0.036055
5 bar two 1.395117 0.716503
6 foo one 0.977119 -1.419242
7 foo three -0.498457 0.912835
Grouping and then applying the sum() function to the resulting
groups.
In [65]: df.groupby('A').sum().execute()
Out[65]:
C D
A
bar 0.311030 -0.356182
foo 1.146954 1.568060
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 -2.435075 -1.399957
three 1.350988 0.327271
two 1.395117 0.716503
foo one 1.813965 -1.186018
three -0.498457 0.912835
two -0.168554 1.841242
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 0x7f26dee51810>
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 -1.027586 0.481455 0.665315 0.424971
1 2000-01-02 -1.025052 0.794528 0.179936 1.051374
2 2000-01-03 -0.466694 -0.093162 1.038509 0.880496
3 2000-01-04 -2.529604 0.907036 2.880652 0.174588
4 2000-01-05 -3.660128 0.464360 4.181216 -1.662126
.. ... ... ... ... ...
995 2002-09-22 -1.573871 -37.825188 7.725830 -34.907312
996 2002-09-23 -0.833625 -38.069050 7.503748 -35.848106
997 2002-09-24 -2.195992 -38.981021 5.207587 -37.535640
998 2002-09-25 -2.365478 -38.622382 5.315007 -37.387363
999 2002-09-26 -3.233688 -38.618008 5.371838 -36.989378
[1000 rows x 5 columns]