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.tensor as mt In [2]: import mars.dataframe as md
Creating a Series by passing a list of values, letting it create a default integer index:
Series
In [3]: s = md.Series([1, 3, 5, mt.nan, 6, 8]) In [4]: s.execute() Out[4]: 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:
DataFrame
In [5]: dates = md.date_range('20130101', periods=6) In [6]: dates.execute() Out[6]: 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 [7]: df = md.DataFrame(mt.random.randn(6, 4), index=dates, columns=list('ABCD')) In [8]: df.execute() Out[8]: A B C D 2013-01-01 0.557885 0.470411 -0.873449 -0.070733 2013-01-02 0.675828 0.901187 -0.453088 -0.104001 2013-01-03 1.282714 0.658673 -0.552264 0.901889 2013-01-04 1.187872 0.506431 0.740055 2.086557 2013-01-05 0.565737 1.137972 -0.747667 1.917502 2013-01-06 0.653736 0.610252 -0.368207 -0.337543
Creating a DataFrame by passing a dict of objects that can be converted to series-like.
In [9]: 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 [10]: df2.execute() Out[10]: 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 [11]: df2.dtypes Out[11]: A float64 B datetime64[ns] C float32 D int32 E object dtype: object
Here is how to view the top and bottom rows of the frame:
In [12]: df.head().execute() Out[12]: A B C D 2013-01-01 0.557885 0.470411 -0.873449 -0.070733 2013-01-02 0.675828 0.901187 -0.453088 -0.104001 2013-01-03 1.282714 0.658673 -0.552264 0.901889 2013-01-04 1.187872 0.506431 0.740055 2.086557 2013-01-05 0.565737 1.137972 -0.747667 1.917502 In [13]: df.tail(3).execute() Out[13]: A B C D 2013-01-04 1.187872 0.506431 0.740055 2.086557 2013-01-05 0.565737 1.137972 -0.747667 1.917502 2013-01-06 0.653736 0.610252 -0.368207 -0.337543
Display the index, columns:
In [14]: df.index.execute() Out[14]: 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 [15]: df.columns.execute() Out[15]: 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.
DataFrame.to_tensor()
object
For df, our DataFrame of all floating-point values, DataFrame.to_tensor() is fast and doesn’t require copying data.
df
In [16]: df.to_tensor().execute() Out[16]: array([[ 0.55788455, 0.47041061, -0.87344889, -0.0707333 ], [ 0.67582823, 0.90118741, -0.45308801, -0.10400123], [ 1.28271368, 0.65867339, -0.55226379, 0.90188851], [ 1.18787188, 0.50643116, 0.74005464, 2.08655746], [ 0.56573706, 1.13797218, -0.74766677, 1.91750157], [ 0.65373621, 0.61025222, -0.36820704, -0.3375432 ]])
For df2, the DataFrame with multiple dtypes, DataFrame.to_tensor() is relatively expensive.
df2
In [17]: df2.to_tensor().execute() Out[17]: 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:
describe()
In [18]: df.describe().execute() Out[18]: A B C D count 6.000000 6.000000 6.000000 6.000000 mean 0.820629 0.714154 -0.375770 0.732278 std 0.325949 0.257308 0.577623 1.072968 min 0.557885 0.470411 -0.873449 -0.337543 25% 0.587737 0.532386 -0.698816 -0.095684 50% 0.664782 0.634463 -0.502676 0.415578 75% 1.059861 0.840559 -0.389427 1.663598 max 1.282714 1.137972 0.740055 2.086557
Sorting by an axis:
In [19]: df.sort_index(axis=1, ascending=False).execute() Out[19]: D C B A 2013-01-01 -0.070733 -0.873449 0.470411 0.557885 2013-01-02 -0.104001 -0.453088 0.901187 0.675828 2013-01-03 0.901889 -0.552264 0.658673 1.282714 2013-01-04 2.086557 0.740055 0.506431 1.187872 2013-01-05 1.917502 -0.747667 1.137972 0.565737 2013-01-06 -0.337543 -0.368207 0.610252 0.653736
Sorting by values:
In [20]: df.sort_values(by='B').execute() Out[20]: A B C D 2013-01-01 0.557885 0.470411 -0.873449 -0.070733 2013-01-04 1.187872 0.506431 0.740055 2.086557 2013-01-06 0.653736 0.610252 -0.368207 -0.337543 2013-01-03 1.282714 0.658673 -0.552264 0.901889 2013-01-02 0.675828 0.901187 -0.453088 -0.104001 2013-01-05 0.565737 1.137972 -0.747667 1.917502
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.
.at
.iat
.loc
.iloc
Selecting a single column, which yields a Series, equivalent to df.A:
df.A
In [21]: df['A'].execute() Out[21]: 2013-01-01 0.557885 2013-01-02 0.675828 2013-01-03 1.282714 2013-01-04 1.187872 2013-01-05 0.565737 2013-01-06 0.653736 Freq: D, Name: A, dtype: float64
Selecting via [], which slices the rows.
[]
In [22]: df[0:3].execute() Out[22]: A B C D 2013-01-01 0.557885 0.470411 -0.873449 -0.070733 2013-01-02 0.675828 0.901187 -0.453088 -0.104001 2013-01-03 1.282714 0.658673 -0.552264 0.901889 In [23]: df['20130102':'20130104'].execute() Out[23]: A B C D 2013-01-02 0.675828 0.901187 -0.453088 -0.104001 2013-01-03 1.282714 0.658673 -0.552264 0.901889 2013-01-04 1.187872 0.506431 0.740055 2.086557
For getting a cross section using a label:
In [24]: df.loc['20130101'].execute() Out[24]: A 0.557885 B 0.470411 C -0.873449 D -0.070733 Name: 2013-01-01 00:00:00, dtype: float64
Selecting on a multi-axis by label:
In [25]: df.loc[:, ['A', 'B']].execute() Out[25]: A B 2013-01-01 0.557885 0.470411 2013-01-02 0.675828 0.901187 2013-01-03 1.282714 0.658673 2013-01-04 1.187872 0.506431 2013-01-05 0.565737 1.137972 2013-01-06 0.653736 0.610252
Showing label slicing, both endpoints are included:
In [26]: df.loc['20130102':'20130104', ['A', 'B']].execute() Out[26]: A B 2013-01-02 0.675828 0.901187 2013-01-03 1.282714 0.658673 2013-01-04 1.187872 0.506431
Reduction in the dimensions of the returned object:
In [27]: df.loc['20130102', ['A', 'B']].execute() Out[27]: A 0.675828 B 0.901187 Name: 2013-01-02 00:00:00, dtype: float64
For getting a scalar value:
In [28]: df.loc['20130101', 'A'].execute() Out[28]: 0.5578845459804017
For getting fast access to a scalar (equivalent to the prior method):
In [29]: df.at['20130101', 'A'].execute() Out[29]: 0.5578845459804017
Select via the position of the passed integers:
In [30]: df.iloc[3].execute() Out[30]: A 1.187872 B 0.506431 C 0.740055 D 2.086557 Name: 2013-01-04 00:00:00, dtype: float64
By integer slices, acting similar to numpy/python:
In [31]: df.iloc[3:5, 0:2].execute() Out[31]: A B 2013-01-04 1.187872 0.506431 2013-01-05 0.565737 1.137972
By lists of integer position locations, similar to the numpy/python style:
In [32]: df.iloc[[1, 2, 4], [0, 2]].execute() Out[32]: A C 2013-01-02 0.675828 -0.453088 2013-01-03 1.282714 -0.552264 2013-01-05 0.565737 -0.747667
For slicing rows explicitly:
In [33]: df.iloc[1:3, :].execute() Out[33]: A B C D 2013-01-02 0.675828 0.901187 -0.453088 -0.104001 2013-01-03 1.282714 0.658673 -0.552264 0.901889
For slicing columns explicitly:
In [34]: df.iloc[:, 1:3].execute() Out[34]: B C 2013-01-01 0.470411 -0.873449 2013-01-02 0.901187 -0.453088 2013-01-03 0.658673 -0.552264 2013-01-04 0.506431 0.740055 2013-01-05 1.137972 -0.747667 2013-01-06 0.610252 -0.368207
For getting a value explicitly:
In [35]: df.iloc[1, 1].execute() Out[35]: 0.9011874124512463
In [36]: df.iat[1, 1].execute() Out[36]: 0.9011874124512463
Using a single column’s values to select data.
In [37]: df[df['A'] > 0].execute() Out[37]: A B C D 2013-01-01 0.557885 0.470411 -0.873449 -0.070733 2013-01-02 0.675828 0.901187 -0.453088 -0.104001 2013-01-03 1.282714 0.658673 -0.552264 0.901889 2013-01-04 1.187872 0.506431 0.740055 2.086557 2013-01-05 0.565737 1.137972 -0.747667 1.917502 2013-01-06 0.653736 0.610252 -0.368207 -0.337543
Selecting values from a DataFrame where a boolean condition is met.
In [38]: df[df > 0].execute() Out[38]: A B C D 2013-01-01 0.557885 0.470411 NaN NaN 2013-01-02 0.675828 0.901187 NaN NaN 2013-01-03 1.282714 0.658673 NaN 0.901889 2013-01-04 1.187872 0.506431 0.740055 2.086557 2013-01-05 0.565737 1.137972 NaN 1.917502 2013-01-06 0.653736 0.610252 NaN NaN
Operations in general exclude missing data.
Performing a descriptive statistic:
In [39]: df.mean().execute() Out[39]: A 0.820629 B 0.714154 C -0.375770 D 0.732278 dtype: float64
Same operation on the other axis:
In [40]: df.mean(1).execute() Out[40]: 2013-01-01 0.021028 2013-01-02 0.254982 2013-01-03 0.572753 2013-01-04 1.130229 2013-01-05 0.718386 2013-01-06 0.139560 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 [41]: s = md.Series([1, 3, 5, mt.nan, 6, 8], index=dates).shift(2) In [42]: s.execute() Out[42]: 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 [43]: df.sub(s, axis='index').execute() Out[43]: A B C D 2013-01-01 NaN NaN NaN NaN 2013-01-02 NaN NaN NaN NaN 2013-01-03 0.282714 -0.341327 -1.552264 -0.098111 2013-01-04 -1.812128 -2.493569 -2.259945 -0.913443 2013-01-05 -4.434263 -3.862028 -5.747667 -3.082498 2013-01-06 NaN NaN NaN NaN
Applying functions to the data:
In [44]: df.apply(lambda x: x.max() - x.min()).execute() Out[44]: A 0.724829 B 0.667562 C 1.613504 D 2.424101 dtype: float64
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 [45]: s = md.Series(['A', 'B', 'C', 'Aaba', 'Baca', mt.nan, 'CABA', 'dog', 'cat']) In [46]: s.str.lower().execute() Out[46]: 0 a 1 b 2 c 3 aaba 4 baca 5 NaN 6 caba 7 dog 8 cat dtype: object
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():
concat()
In [47]: df = md.DataFrame(mt.random.randn(10, 4)) In [48]: df.execute() Out[48]: 0 1 2 3 0 0.303937 0.676581 0.297204 0.338003 1 -1.541509 1.164739 1.170399 0.051969 2 -0.262747 1.534308 -0.556279 -0.101921 3 -1.327300 -0.028588 -0.724176 -1.103875 4 -1.550649 -0.095343 -0.266349 -0.888187 5 -0.042037 1.028927 -0.820438 1.022288 6 0.991051 -0.513336 -1.557178 0.697850 7 1.680323 0.842725 1.755703 0.911513 8 -0.831695 -0.636868 0.676978 0.143900 9 -0.064976 -0.802800 -1.541547 0.833808 # break it into pieces In [49]: pieces = [df[:3], df[3:7], df[7:]] In [50]: md.concat(pieces).execute() Out[50]: 0 1 2 3 0 0.303937 0.676581 0.297204 0.338003 1 -1.541509 1.164739 1.170399 0.051969 2 -0.262747 1.534308 -0.556279 -0.101921 3 -1.327300 -0.028588 -0.724176 -1.103875 4 -1.550649 -0.095343 -0.266349 -0.888187 5 -0.042037 1.028927 -0.820438 1.022288 6 0.991051 -0.513336 -1.557178 0.697850 7 1.680323 0.842725 1.755703 0.911513 8 -0.831695 -0.636868 0.676978 0.143900 9 -0.064976 -0.802800 -1.541547 0.833808
Adding a column to a DataFrame is relatively fast. However, adding a row requires a copy, and may be expensive. We recommend passing a pre-built list of records to the DataFrame constructor instead of building a DataFrame by iteratively appending records to it.
SQL style merges. See the Database style joining section.
In [51]: left = md.DataFrame({'key': ['foo', 'foo'], 'lval': [1, 2]}) In [52]: right = md.DataFrame({'key': ['foo', 'foo'], 'rval': [4, 5]}) In [53]: left.execute() Out[53]: key lval 0 foo 1 1 foo 2 In [54]: right.execute() Out[54]: key rval 0 foo 4 1 foo 5 In [55]: md.merge(left, right, on='key').execute() Out[55]: 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 [56]: left = md.DataFrame({'key': ['foo', 'bar'], 'lval': [1, 2]}) In [57]: right = md.DataFrame({'key': ['foo', 'bar'], 'rval': [4, 5]}) In [58]: left.execute() Out[58]: key lval 0 foo 1 1 bar 2 In [59]: right.execute() Out[59]: key rval 0 foo 4 1 bar 5 In [60]: md.merge(left, right, on='key').execute() Out[60]: key lval rval 0 foo 1 4 1 bar 2 5
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
Splitting the data into groups based on some criteria
Applying a function to each group independently
Combining the results into a data structure
In [61]: 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 [62]: df.execute() Out[62]: A B C D 0 foo one 0.236483 -0.283515 1 bar one 0.778070 -0.065016 2 foo two 0.804944 0.779283 3 bar three -1.039868 -1.075678 4 foo two -0.050011 -1.887501 5 bar two -1.723543 -0.142226 6 foo one -0.542069 -0.297924 7 foo three -0.023324 0.753029
Grouping and then applying the sum() function to the resulting groups.
sum()
In [63]: df.groupby('A').sum().execute() Out[63]: C D A bar -1.985342 -1.282920 foo 0.426024 -0.936627
Grouping by multiple columns forms a hierarchical index, and again we can apply the sum function.
In [64]: df.groupby(['A', 'B']).sum().execute() Out[64]: C D A B foo one -0.305585 -0.581439 two 0.754933 -1.108218 three -0.023324 0.753029 bar one 0.778070 -0.065016 two -1.723543 -0.142226 three -1.039868 -1.075678
We use the standard convention for referencing the matplotlib API:
In [65]: import matplotlib.pyplot as plt In [66]: plt.close('all')
In [67]: ts = md.Series(mt.random.randn(1000), ....: index=md.date_range('1/1/2000', periods=1000)) ....: In [68]: ts = ts.cumsum() In [69]: ts.plot() Out[69]: <AxesSubplot:>
On a DataFrame, the plot() method is a convenience to plot all of the columns with labels:
plot()
In [70]: df = md.DataFrame(mt.random.randn(1000, 4), index=ts.index, ....: columns=['A', 'B', 'C', 'D']) ....: In [71]: df = df.cumsum() In [72]: plt.figure() Out[72]: <Figure size 640x480 with 0 Axes> In [73]: df.plot() Out[73]: <AxesSubplot:> In [74]: plt.legend(loc='best') Out[74]: <matplotlib.legend.Legend at 0x7ff762a3d790>
In [75]: df.to_csv('foo.csv').execute() Out[75]: Empty DataFrame Columns: [] Index: []
Reading from a csv file.
In [76]: md.read_csv('foo.csv').execute() Out[76]: Unnamed: 0 A B C D 0 2000-01-01 0.476736 -0.530359 -0.028306 1.219260 1 2000-01-02 -0.144028 -1.370880 -1.200896 0.045540 2 2000-01-03 0.050505 -0.403787 -0.502491 0.491048 3 2000-01-04 0.765867 -1.120256 -0.812434 -0.595760 4 2000-01-05 2.440408 0.357761 -1.474871 0.838766 .. ... ... ... ... ... 995 2002-09-22 -32.183753 -1.571340 -26.254543 -50.510185 996 2002-09-23 -33.072655 -2.267491 -27.223734 -51.296070 997 2002-09-24 -31.575486 -1.972108 -28.569647 -51.134554 998 2002-09-25 -30.922447 -1.018882 -26.659718 -51.022256 999 2002-09-26 -31.779789 -0.147407 -25.589601 -49.144979 [1000 rows x 5 columns]