juan503949201的个人博客分享 http://blog.sciencenet.cn/u/juan503949201

博文

10 Minutes to pandas

已有 3483 次阅读 2015-1-12 15:57 |个人分类:python|系统分类:科研笔记

This is a short introduction to pandas, geared mainly for new users. You can see more complex recipes in the Cookbook

Customarily, we import as follows

In [1]: importpandasaspdIn [2]: importnumpyasnpIn [3]: importmatplotlib.pyplotasplt
Object Creation

See the Data Structure Intro section

Creating a Series by passing a list of values, letting pandas create a default integer index

In [4]: s=pd.Series([1,3,5,np.nan,6,8])In [5]: sOut[5]: 0     11     32     53   NaN4     65     8dtype: float64

Creating a DataFrame by passing a numpy array, with a datetime index and labeled columns.

In [6]: dates=pd.date_range('20130101',periods=6)In [7]: datesOut[7]: <class 'pandas.tseries.index.DatetimeIndex'>[2013-01-01, ..., 2013-01-06]Length: 6, Freq: D, Timezone: NoneIn [8]: df=pd.DataFrame(np.random.randn(6,4),index=dates,columns=list('ABCD'))In [9]: dfOut[9]:                   A         B         C         D2013-01-01  0.469112 -0.282863 -1.509059 -1.1356322013-01-02  1.212112 -0.173215  0.119209 -1.0442362013-01-03 -0.861849 -2.104569 -0.494929  1.0718042013-01-04  0.721555 -0.706771 -1.039575  0.2718602013-01-05 -0.424972  0.567020  0.276232 -1.0874012013-01-06 -0.673690  0.113648 -1.478427  0.524988

Creating a DataFrame by passing a dict of objects that can be converted to series-like.

In [10]: df2=pd.DataFrame({'A':1.,   ....:'B':pd.Timestamp('20130102'),   ....:'C':pd.Series(1,index=list(range(4)),dtype='float32'),   ....:'D':np.array([3]*4,dtype='int32'),   ....:'E':pd.Categorical(["test","train","test","train"]),   ....:'F':'foo'})   ....:In [11]: df2Out[11]:   A          B  C  D      E    F0  1 2013-01-02  1  3   test  foo1  1 2013-01-02  1  3  train  foo2  1 2013-01-02  1  3   test  foo3  1 2013-01-02  1  3  train  foo

Having specific dtypes

In [12]: df2.dtypesOut[12]: A           float64B    datetime64[ns]C           float32D             int32E          categoryF            objectdtype: object

If you’re using IPython, tab completion for column names (as well as public attributes) is automatically enabled. Here’s a subset of the attributes that will be completed:

In [13]: df2.<TAB>df2.A                  df2.boxplotdf2.abs                df2.Cdf2.add                df2.clipdf2.add_prefix         df2.clip_lowerdf2.add_suffix         df2.clip_upperdf2.align              df2.columnsdf2.all                df2.combinedf2.any                df2.combineAdddf2.append             df2.combine_firstdf2.apply              df2.combineMultdf2.applymap           df2.compounddf2.as_blocks          df2.consolidatedf2.asfreq             df2.convert_objectsdf2.as_matrix          df2.copydf2.astype             df2.corrdf2.at                 df2.corrwithdf2.at_time            df2.countdf2.axes               df2.covdf2.B                  df2.cummaxdf2.between_time       df2.cummindf2.bfill              df2.cumproddf2.blocks             df2.cumsumdf2.bool               df2.D

As you can see, the columns A, B, C, and D are automatically tab completed. E is there as well; the rest of the attributes have been truncated for brevity.

Viewing Data

See the Basics section

See the top & bottom rows of the frame

In [14]: df.head()Out[14]:                   A         B         C         D2013-01-01  0.469112 -0.282863 -1.509059 -1.1356322013-01-02  1.212112 -0.173215  0.119209 -1.0442362013-01-03 -0.861849 -2.104569 -0.494929  1.0718042013-01-04  0.721555 -0.706771 -1.039575  0.2718602013-01-05 -0.424972  0.567020  0.276232 -1.087401In [15]: df.tail(3)Out[15]:                   A         B         C         D2013-01-04  0.721555 -0.706771 -1.039575  0.2718602013-01-05 -0.424972  0.567020  0.276232 -1.0874012013-01-06 -0.673690  0.113648 -1.478427  0.524988

Display the index,columns, and the underlying numpy data

In [16]: df.indexOut[16]: <class 'pandas.tseries.index.DatetimeIndex'>[2013-01-01, ..., 2013-01-06]Length: 6, Freq: D, Timezone: NoneIn [17]: df.columnsOut[17]: Index([u'A',u'B',u'C',u'D'],dtype='object')In [18]: df.valuesOut[18]: array([[ 0.4691, -0.2829, -1.5091, -1.1356],       [ 1.2121, -0.1732,  0.1192, -1.0442],       [-0.8618, -2.1046, -0.4949,  1.0718],       [ 0.7216, -0.7068, -1.0396,  0.2719],       [-0.425 ,  0.567 ,  0.2762, -1.0874],       [-0.6737,  0.1136, -1.4784,  0.525 ]])

Describe shows a quick statistic summary of your data

In [19]: df.describe()Out[19]:              A         B         C         Dcount  6.000000  6.000000  6.000000  6.000000mean   0.073711 -0.431125 -0.687758 -0.233103std    0.843157  0.922818  0.779887  0.973118min   -0.861849 -2.104569 -1.509059 -1.13563225%   -0.611510 -0.600794 -1.368714 -1.07661050%    0.022070 -0.228039 -0.767252 -0.38618875%    0.658444  0.041933 -0.034326  0.461706max    1.212112  0.567020  0.276232  1.071804

Transposing your data

In [20]: df.TOut[20]:   2013-01-01  2013-01-02  2013-01-03  2013-01-04  2013-01-05  2013-01-06A    0.469112    1.212112   -0.861849    0.721555   -0.424972   -0.673690B   -0.282863   -0.173215   -2.104569   -0.706771    0.567020    0.113648C   -1.509059    0.119209   -0.494929   -1.039575    0.276232   -1.478427D   -1.135632   -1.044236    1.071804    0.271860   -1.087401    0.524988

Sorting by an axis

In [21]: df.sort_index(axis=1,ascending=False)Out[21]:                   D         C         B         A2013-01-01 -1.135632 -1.509059 -0.282863  0.4691122013-01-02 -1.044236  0.119209 -0.173215  1.2121122013-01-03  1.071804 -0.494929 -2.104569 -0.8618492013-01-04  0.271860 -1.039575 -0.706771  0.7215552013-01-05 -1.087401  0.276232  0.567020 -0.4249722013-01-06  0.524988 -1.478427  0.113648 -0.673690

Sorting by values

In [22]: df.sort(columns='B')Out[22]:                   A         B         C         D2013-01-03 -0.861849 -2.104569 -0.494929  1.0718042013-01-04  0.721555 -0.706771 -1.039575  0.2718602013-01-01  0.469112 -0.282863 -1.509059 -1.1356322013-01-02  1.212112 -0.173215  0.119209 -1.0442362013-01-06 -0.673690  0.113648 -1.478427  0.5249882013-01-05 -0.424972  0.567020  0.276232 -1.087401
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 pandas data access methods, .at, .iat, .loc, .iloc and .ix.

See the indexing documentation Indexing and Selecing Data and MultiIndex / Advanced Indexing

Getting

Selecting a single column, which yields a Series, equivalent to df.A

In [23]: df['A']Out[23]: 2013-01-01    0.4691122013-01-02    1.2121122013-01-03   -0.8618492013-01-04    0.7215552013-01-05   -0.4249722013-01-06   -0.673690Freq: D, Name: A, dtype: float64

Selecting via [], which slices the rows.

In [24]: df[0:3]Out[24]:                   A         B         C         D2013-01-01  0.469112 -0.282863 -1.509059 -1.1356322013-01-02  1.212112 -0.173215  0.119209 -1.0442362013-01-03 -0.861849 -2.104569 -0.494929  1.071804In [25]: df['20130102':'20130104']Out[25]:                   A         B         C         D2013-01-02  1.212112 -0.173215  0.119209 -1.0442362013-01-03 -0.861849 -2.104569 -0.494929  1.0718042013-01-04  0.721555 -0.706771 -1.039575  0.271860
Selection by Label

See more in Selection by Label

For getting a cross section using a label

In [26]: df.loc[dates[0]]Out[26]: A    0.469112B   -0.282863C   -1.509059D   -1.135632Name: 2013-01-01 00:00:00, dtype: float64

Selecting on a multi-axis by label

In [27]: df.loc[:,['A','B']]Out[27]:                   A         B2013-01-01  0.469112 -0.2828632013-01-02  1.212112 -0.1732152013-01-03 -0.861849 -2.1045692013-01-04  0.721555 -0.7067712013-01-05 -0.424972  0.5670202013-01-06 -0.673690  0.113648

Showing label slicing, both endpoints are included

In [28]: df.loc['20130102':'20130104',['A','B']]Out[28]:                   A         B2013-01-02  1.212112 -0.1732152013-01-03 -0.861849 -2.1045692013-01-04  0.721555 -0.706771

Reduction in the dimensions of the returned object

In [29]: df.loc['20130102',['A','B']]Out[29]: A    1.212112B   -0.173215Name: 2013-01-02 00:00:00, dtype: float64

For getting a scalar value

In [30]: df.loc[dates[0],'A']Out[30]: 0.46911229990718628

For getting fast access to a scalar (equiv to the prior method)

In [31]: df.at[dates[0],'A']Out[31]: 0.46911229990718628
Selection by Position

See more in Selection by Position

Select via the position of the passed integers

In [32]: df.iloc[3]Out[32]: A    0.721555B   -0.706771C   -1.039575D    0.271860Name: 2013-01-04 00:00:00, dtype: float64

By integer slices, acting similar to numpy/python

In [33]: df.iloc[3:5,0:2]Out[33]:                   A         B2013-01-04  0.721555 -0.7067712013-01-05 -0.424972  0.567020

By lists of integer position locations, similar to the numpy/python style

In [34]: df.iloc[[1,2,4],[0,2]]Out[34]:                   A         C2013-01-02  1.212112  0.1192092013-01-03 -0.861849 -0.4949292013-01-05 -0.424972  0.276232

For slicing rows explicitly

In [35]: df.iloc[1:3,:]Out[35]:                   A         B         C         D2013-01-02  1.212112 -0.173215  0.119209 -1.0442362013-01-03 -0.861849 -2.104569 -0.494929  1.071804

For slicing columns explicitly

In [36]: df.iloc[:,1:3]Out[36]:                   B         C2013-01-01 -0.282863 -1.5090592013-01-02 -0.173215  0.1192092013-01-03 -2.104569 -0.4949292013-01-04 -0.706771 -1.0395752013-01-05  0.567020  0.2762322013-01-06  0.113648 -1.478427

For getting a value explicitly

In [37]: df.iloc[1,1]Out[37]: -0.17321464905330861

For getting fast access to a scalar (equiv to the prior method)

In [38]: df.iat[1,1]Out[38]: -0.17321464905330861
Boolean Indexing

Using a single column’s values to select data.

In [39]: df[df.A>0]Out[39]:                   A         B         C         D2013-01-01  0.469112 -0.282863 -1.509059 -1.1356322013-01-02  1.212112 -0.173215  0.119209 -1.0442362013-01-04  0.721555 -0.706771 -1.039575  0.271860

A where operation for getting.

In [40]: df[df>0]Out[40]:                   A         B         C         D2013-01-01  0.469112       NaN       NaN       NaN2013-01-02  1.212112       NaN  0.119209       NaN2013-01-03       NaN       NaN       NaN  1.0718042013-01-04  0.721555       NaN       NaN  0.2718602013-01-05       NaN  0.567020  0.276232       NaN2013-01-06       NaN  0.113648       NaN  0.524988

Using the isin() method for filtering:

In [41]: df2=df.copy()In [42]: df2['E']=['one','one','two','three','four','three']In [43]: df2Out[43]:                   A         B         C         D      E2013-01-01  0.469112 -0.282863 -1.509059 -1.135632    one2013-01-02  1.212112 -0.173215  0.119209 -1.044236    one2013-01-03 -0.861849 -2.104569 -0.494929  1.071804    two2013-01-04  0.721555 -0.706771 -1.039575  0.271860  three2013-01-05 -0.424972  0.567020  0.276232 -1.087401   four2013-01-06 -0.673690  0.113648 -1.478427  0.524988  threeIn [44]: df2[df2['E'].isin(['two','four'])]Out[44]:                   A         B         C         D     E2013-01-03 -0.861849 -2.104569 -0.494929  1.071804   two2013-01-05 -0.424972  0.567020  0.276232 -1.087401  four
Setting

Setting a new column automatically aligns the data by the indexes

In [45]: s1=pd.Series([1,2,3,4,5,6],index=pd.date_range('20130102',periods=6))In [46]: s1Out[46]: 2013-01-02    12013-01-03    22013-01-04    32013-01-05    42013-01-06    52013-01-07    6Freq: D, dtype: int64In [47]: df['F']=s1

Setting values by label

In [48]: df.at[dates[0],'A']=0

Setting values by position

In [49]: df.iat[0,1]=0

Setting by assigning with a numpy array

In [50]: df.loc[:,'D']=np.array([5]*len(df))

The result of the prior setting operations

In [51]: dfOut[51]:                   A         B         C  D   F2013-01-01  0.000000  0.000000 -1.509059  5 NaN2013-01-02  1.212112 -0.173215  0.119209  5   12013-01-03 -0.861849 -2.104569 -0.494929  5   22013-01-04  0.721555 -0.706771 -1.039575  5   32013-01-05 -0.424972  0.567020  0.276232  5   42013-01-06 -0.673690  0.113648 -1.478427  5   5

A where operation with setting.

In [52]: df2=df.copy()In [53]: df2[df2>0]=-df2In [54]: df2Out[54]:                   A         B         C  D   F2013-01-01  0.000000  0.000000 -1.509059 -5 NaN2013-01-02 -1.212112 -0.173215 -0.119209 -5  -12013-01-03 -0.861849 -2.104569 -0.494929 -5  -22013-01-04 -0.721555 -0.706771 -1.039575 -5  -32013-01-05 -0.424972 -0.567020 -0.276232 -5  -42013-01-06 -0.673690 -0.113648 -1.478427 -5  -5
Missing Data

pandas primarily uses the value np.nan to represent missing data. It is by default not included in computations. See the Missing Data section

Reindexing allows you to change/add/delete the index on a specified axis. This returns a copy of the data.

In [55]: df1=df.reindex(index=dates[0:4],columns=list(df.columns)+['E'])In [56]: df1.loc[dates[0]:dates[1],'E']=1In [57]: df1Out[57]:                   A         B         C  D   F   E2013-01-01  0.000000  0.000000 -1.509059  5 NaN   12013-01-02  1.212112 -0.173215  0.119209  5   1   12013-01-03 -0.861849 -2.104569 -0.494929  5   2 NaN2013-01-04  0.721555 -0.706771 -1.039575  5   3 NaN

To drop any rows that have missing data.

In [58]: df1.dropna(how='any')Out[58]:                   A         B         C  D  F  E2013-01-02  1.212112 -0.173215  0.119209  5  1  1

Filling missing data

In [59]: df1.fillna(value=5)Out[59]:                   A         B         C  D  F  E2013-01-01  0.000000  0.000000 -1.509059  5  5  12013-01-02  1.212112 -0.173215  0.119209  5  1  12013-01-03 -0.861849 -2.104569 -0.494929  5  2  52013-01-04  0.721555 -0.706771 -1.039575  5  3  5

To get the boolean mask where values are nan

In [60]: pd.isnull(df1)Out[60]:                A      B      C      D      F      E2013-01-01  False  False  False  False   True  False2013-01-02  False  False  False  False  False  False2013-01-03  False  False  False  False  False   True2013-01-04  False  False  False  False  False   True
Operations

See the Basic section on Binary Ops

Stats

Operations in general exclude missing data.

Performing a descriptive statistic

In [61]: df.mean()Out[61]: A   -0.004474B   -0.383981C   -0.687758D    5.000000F    3.000000dtype: float64

Same operation on the other axis

In [62]: df.mean(1)Out[62]: 2013-01-01    0.8727352013-01-02    1.4316212013-01-03    0.7077312013-01-04    1.3950422013-01-05    1.8836562013-01-06    1.592306Freq: D, dtype: float64

Operating with objects that have different dimensionality and need alignment. In addition, pandas automatically broadcasts along the specified dimension.

In [63]: s=pd.Series([1,3,5,np.nan,6,8],index=dates).shift(2)In [64]: sOut[64]: 2013-01-01   NaN2013-01-02   NaN2013-01-03     12013-01-04     32013-01-05     52013-01-06   NaNFreq: D, dtype: float64In [65]: df.sub(s,axis='index')Out[65]:                   A         B         C   D   F2013-01-01       NaN       NaN       NaN NaN NaN2013-01-02       NaN       NaN       NaN NaN NaN2013-01-03 -1.861849 -3.104569 -1.494929   4   12013-01-04 -2.278445 -3.706771 -4.039575   2   02013-01-05 -5.424972 -4.432980 -4.723768   0  -12013-01-06       NaN       NaN       NaN NaN NaN
Apply

Applying functions to the data

In [66]: df.apply(np.cumsum)Out[66]:                   A         B         C   D   F2013-01-01  0.000000  0.000000 -1.509059   5 NaN2013-01-02  1.212112 -0.173215 -1.389850  10   12013-01-03  0.350263 -2.277784 -1.884779  15   32013-01-04  1.071818 -2.984555 -2.924354  20   62013-01-05  0.646846 -2.417535 -2.648122  25  102013-01-06 -0.026844 -2.303886 -4.126549  30  15In [67]: df.apply(lambdax:x.max()-x.min())Out[67]: A    2.073961B    2.671590C    1.785291D    0.000000F    4.000000dtype: float64
Histogramming

See more at Histogramming and Discretization

In [68]: s=pd.Series(np.random.randint(0,7,size=10))In [69]: sOut[69]: 0    41    22    13    24    65    46    47    68    49    4dtype: int32In [70]: s.value_counts()Out[70]: 4    56    22    21    1dtype: int64
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 strgenerally uses regular expressions by default (and in some cases always uses them). See more atVectorized String Methods.

In [71]: s=pd.Series(['A','B','C','Aaba','Baca',np.nan,'CABA','dog','cat'])In [72]: s.str.lower()Out[72]: 0       a1       b2       c3    aaba4    baca5     NaN6    caba7     dog8     catdtype: object
Merge
Concat

pandas provides various facilities for easily combining together Series, DataFrame, and Panel objects with various kinds of set logic for the indexes and relational algebra functionality in the case of join / merge-type operations.

See the Merging section

Concatenating pandas objects together

In [73]: df=pd.DataFrame(np.random.randn(10,4))In [74]: dfOut[74]:          0         1         2         30 -0.548702  1.467327 -1.015962 -0.4830751  1.637550 -1.217659 -0.291519 -1.7455052 -0.263952  0.991460 -0.919069  0.2660463 -0.709661  1.669052  1.037882 -1.7057754 -0.919854 -0.042379  1.247642 -0.0099205  0.290213  0.495767  0.362949  1.5481066 -1.131345 -0.089329  0.337863 -0.9458677 -0.932132  1.956030  0.017587 -0.0166928 -0.575247  0.254161 -1.143704  0.2158979  1.193555 -0.077118 -0.408530 -0.862495# break it into piecesIn [75]: pieces=[df[:3],df[3:7],df[7:]]In [76]: pd.concat(pieces)Out[76]:          0         1         2         30 -0.548702  1.467327 -1.015962 -0.4830751  1.637550 -1.217659 -0.291519 -1.7455052 -0.263952  0.991460 -0.919069  0.2660463 -0.709661  1.669052  1.037882 -1.7057754 -0.919854 -0.042379  1.247642 -0.0099205  0.290213  0.495767  0.362949  1.5481066 -1.131345 -0.089329  0.337863 -0.9458677 -0.932132  1.956030  0.017587 -0.0166928 -0.575247  0.254161 -1.143704  0.2158979  1.193555 -0.077118 -0.408530 -0.862495
Join

SQL style merges. See the Database style joining

In [77]: left=pd.DataFrame({'key':['foo','foo'],'lval':[1,2]})In [78]: right=pd.DataFrame({'key':['foo','foo'],'rval':[4,5]})In [79]: leftOut[79]:   key  lval0  foo     11  foo     2In [80]: rightOut[80]:   key  rval0  foo     41  foo     5In [81]: pd.merge(left,right,on='key')Out[81]:   key  lval  rval0  foo     1     41  foo     1     52  foo     2     43  foo     2     5
Append

Append rows to a dataframe. See the Appending

In [82]: df=pd.DataFrame(np.random.randn(8,4),columns=['A','B','C','D'])In [83]: dfOut[83]:          A         B         C         D0  1.346061  1.511763  1.627081 -0.9905821 -0.441652  1.211526  0.268520  0.0245802 -1.577585  0.396823 -0.105381 -0.5325323  1.453749  1.208843 -0.080952 -0.2646104 -0.727965 -0.589346  0.339969 -0.6932055 -0.339355  0.593616  0.884345  1.5914316  0.141809  0.220390  0.435589  0.1924517 -0.096701  0.803351  1.715071 -0.708758In [84]: s=df.iloc[3]In [85]: df.append(s,ignore_index=True)Out[85]:          A         B         C         D0  1.346061  1.511763  1.627081 -0.9905821 -0.441652  1.211526  0.268520  0.0245802 -1.577585  0.396823 -0.105381 -0.5325323  1.453749  1.208843 -0.080952 -0.2646104 -0.727965 -0.589346  0.339969 -0.6932055 -0.339355  0.593616  0.884345  1.5914316  0.141809  0.220390  0.435589  0.1924517 -0.096701  0.803351  1.715071 -0.7087588  1.453749  1.208843 -0.080952 -0.264610
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

See the Grouping section

In [86]: 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 [87]: dfOut[87]:     A      B         C         D0  foo    one -1.202872 -0.0552241  bar    one -1.814470  2.3959852  foo    two  1.018601  1.5528253  bar  three -0.595447  0.1665994  foo    two  1.395433  0.0476095  bar    two -0.392670 -0.1364736  foo    one  0.007207 -0.5617577  foo  three  1.928123 -1.623033

Grouping and then applying a function sum to the resulting groups.

In [88]: df.groupby('A').sum()Out[88]:            C        DA                     bar -2.802588  2.42611foo  3.146492 -0.63958

Grouping by multiple columns forms a hierarchical index, which we then apply the function.

In [89]: df.groupby(['A','B']).sum()Out[89]:                  C         DA   B                        bar one   -1.814470  2.395985    three -0.595447  0.166599    two   -0.392670 -0.136473foo one   -1.195665 -0.616981    three  1.928123 -1.623033    two    2.414034  1.600434
Reshaping

See the sections on Hierarchical Indexing and Reshaping.

Stack
In [90]: tuples=list(zip(*[['bar','bar','baz','baz',   ....:'foo','foo','qux','qux'],   ....:['one','two','one','two',   ....:'one','two','one','two']]))   ....:In [91]: index=pd.MultiIndex.from_tuples(tuples,names=['first','second'])In [92]: df=pd.DataFrame(np.random.randn(8,2),index=index,columns=['A','B'])In [93]: df2=df[:4]In [94]: df2Out[94]:                     A         Bfirst second                    bar   one     0.029399 -0.542108      two     0.282696 -0.087302baz   one    -1.575170  1.771208      two     0.816482  1.100230

The stack function “compresses” a level in the DataFrame’s columns.

In [95]: stacked=df2.stack()In [96]: stackedOut[96]: first  second   bar    one     A    0.029399               B   -0.542108       two     A    0.282696               B   -0.087302baz    one     A   -1.575170               B    1.771208       two     A    0.816482               B    1.100230dtype: float64

With a “stacked” DataFrame or Series (having a MultiIndex as the index), the inverse operation of stackis unstack, which by default unstacks the last level:

In [97]: stacked.unstack()Out[97]:                     A         Bfirst second                    bar   one     0.029399 -0.542108      two     0.282696 -0.087302baz   one    -1.575170  1.771208      two     0.816482  1.100230In [98]: stacked.unstack(1)Out[98]: second        one       twofirst                      bar   A  0.029399  0.282696      B -0.542108 -0.087302baz   A -1.575170  0.816482      B  1.771208  1.100230In [99]: stacked.unstack(0)Out[99]: first          bar       bazsecond                      one    A  0.029399 -1.575170       B -0.542108  1.771208two    A  0.282696  0.816482       B -0.087302  1.100230
Pivot Tables

See the section on Pivot Tables.

In [100]: df=pd.DataFrame({'A':['one','one','two','three']*3,   .....:'B':['A','B','C']*4,   .....:'C':['foo','foo','foo','bar','bar','bar']*2,   .....:'D':np.random.randn(12),   .....:'E':np.random.randn(12)})   .....:In [101]: dfOut[101]:        A  B    C         D         E0     one  A  foo  1.418757 -0.1796661     one  B  foo -1.879024  1.2918362     two  C  foo  0.536826 -0.0096143   three  A  bar  1.006160  0.3921494     one  B  bar -0.029716  0.2645995     one  C  bar -1.146178 -0.0574096     two  A  foo  0.100900 -1.4256387   three  B  foo -1.035018  1.0240988     one  C  foo  0.314665 -0.1060629     one  A  bar -0.773723  1.82437510    two  B  bar -1.170653  0.59597411  three  C  bar  0.648740  1.167115

We can produce pivot tables from this data very easily:

In [102]: pd.pivot_table(df,values='D',index=['A','B'],columns=['C'])Out[102]: C             bar       fooA     B                    one   A -0.773723  1.418757      B -0.029716 -1.879024      C -1.146178  0.314665three A  1.006160       NaN      B       NaN -1.035018      C  0.648740       NaNtwo   A       NaN  0.100900      B -1.170653       NaN      C       NaN  0.536826
Time Series

pandas has simple, powerful, and efficient functionality for performing resampling operations during frequency conversion (e.g., converting secondly data into 5-minutely data). This is extremely common in, but not limited to, financial applications. See the Time Series section

In [103]: rng=pd.date_range('1/1/2012',periods=100,freq='S')In [104]: ts=pd.Series(np.random.randint(0,500,len(rng)),index=rng)In [105]: ts.resample('5Min',how='sum')Out[105]: 2012-01-01    25083Freq: 5T, dtype: int32

Time zone representation

In [106]: rng=pd.date_range('3/6/2012 00:00',periods=5,freq='D')In [107]: ts=pd.Series(np.random.randn(len(rng)),rng)In [108]: tsOut[108]: 2012-03-06    0.4640002012-03-07    0.2273712012-03-08   -0.4969222012-03-09    0.3063892012-03-10   -2.290613Freq: D, dtype: float64In [109]: ts_utc=ts.tz_localize('UTC')In [110]: ts_utcOut[110]: 2012-03-06 00:00:00+00:00    0.4640002012-03-07 00:00:00+00:00    0.2273712012-03-08 00:00:00+00:00   -0.4969222012-03-09 00:00:00+00:00    0.3063892012-03-10 00:00:00+00:00   -2.290613Freq: D, dtype: float64

Convert to another time zone

In [111]: ts_utc.tz_convert('US/Eastern')Out[111]: 2012-03-05 19:00:00-05:00    0.4640002012-03-06 19:00:00-05:00    0.2273712012-03-07 19:00:00-05:00   -0.4969222012-03-08 19:00:00-05:00    0.3063892012-03-09 19:00:00-05:00   -2.290613Freq: D, dtype: float64

Converting between time span representations

In [112]: rng=pd.date_range('1/1/2012',periods=5,freq='M')In [113]: ts=pd.Series(np.random.randn(len(rng)),index=rng)In [114]: tsOut[114]: 2012-01-31   -1.1346232012-02-29   -1.5618192012-03-31   -0.2608382012-04-30    0.2819572012-05-31    1.523962Freq: M, dtype: float64In [115]: ps=ts.to_period()In [116]: psOut[116]: 2012-01   -1.1346232012-02   -1.5618192012-03   -0.2608382012-04    0.2819572012-05    1.523962Freq: M, dtype: float64In [117]: ps.to_timestamp()Out[117]: 2012-01-01   -1.1346232012-02-01   -1.5618192012-03-01   -0.2608382012-04-01    0.2819572012-05-01    1.523962Freq: MS, dtype: float64

Converting between period and timestamp enables some convenient arithmetic functions to be used. In the following example, we convert a quarterly frequency with year ending in November to 9am of the end of the month following the quarter end:

In [118]: prng=pd.period_range('1990Q1','2000Q4',freq='Q-NOV')In [119]: ts=pd.Series(np.random.randn(len(prng)),prng)In [120]: ts.index=(prng.asfreq('M','e')+1).asfreq('H','s')+9In [121]: ts.head()Out[121]: 1990-03-01 09:00   -0.9029371990-06-01 09:00    0.0681591990-09-01 09:00   -0.0578731990-12-01 09:00   -0.3682041991-03-01 09:00   -1.144073Freq: H, dtype: float64
Categoricals

Since version 0.15, pandas can include categorical data in a DataFrame. For full docs, see thecategorical introduction and the API documentation.

In [122]: df=pd.DataFrame({"id":[1,2,3,4,5,6],"raw_grade":['a','b','b','a','a','e']})

Convert the raw grades to a categorical data type.

In [123]: df["grade"]=df["raw_grade"].astype("category")In [124]: df["grade"]Out[124]: 0    a1    b2    b3    a4    a5    eName: grade, dtype: categoryCategories (3, object): [a < b < e]

Rename the categories to more meaningful names (assigning to Series.cat.categories is inplace!)

In [125]: df["grade"].cat.categories=["very good","good","very bad"]

Reorder the categories and simultaneously add the missing categories (methods under Series .catreturn a new Series per default).

In [126]: df["grade"]=df["grade"].cat.set_categories(["very bad","bad","medium","good","very good"])In [127]: df["grade"]Out[127]: 0    very good1         good2         good3    very good4    very good5     very badName: grade, dtype: categoryCategories (5, object): [very bad < bad < medium < good < very good]

Sorting is per order in the categories, not lexical order.

In [128]: df.sort("grade")Out[128]:   id raw_grade      grade5   6         e   very bad1   2         b       good2   3         b       good0   1         a  very good3   4         a  very good4   5         a  very good

Grouping by a categorical column shows also empty categories.

In [129]: df.groupby("grade").size()Out[129]: gradevery bad      1bad         NaNmedium      NaNgood          2very good     3dtype: float64
Plotting

Plotting docs.

In [130]: ts=pd.Series(np.random.randn(1000),index=pd.date_range('1/1/2000',periods=1000))In [131]: ts=ts.cumsum()In [132]: ts.plot()Out[132]: <matplotlib.axes._subplots.AxesSubplotat0xaf3e40ac>
_images/series_plot_basic.png

On DataFrame, plot is a convenience to plot all of the columns with labels:

In [133]: df=pd.DataFrame(np.random.randn(1000,4),index=ts.index,   .....:columns=['A','B','C','D'])   .....:In [134]: df=df.cumsum()In [135]: plt.figure();df.plot();plt.legend(loc='best')Out[135]: <matplotlib.legend.Legendat0xaf34476c>
_images/frame_plot_basic.png
Getting Data In/Out
CSV

Writing to a csv file

In [136]: df.to_csv('foo.csv')

Reading from a csv file

In [137]: pd.read_csv('foo.csv')Out[137]:     Unnamed: 0          A          B         C          D0    2000-01-01   0.266457  -0.399641 -0.219582   1.1868601    2000-01-02  -1.170732  -0.345873  1.653061  -0.2829532    2000-01-03  -1.734933   0.530468  2.060811  -0.5155363    2000-01-04  -1.555121   1.452620  0.239859  -1.1568964    2000-01-05   0.578117   0.511371  0.103552  -2.4282025    2000-01-06   0.478344   0.449933 -0.741620  -1.9624096    2000-01-07   1.235339  -0.091757 -1.543861  -1.084753..          ...        ...        ...       ...        ...993  2002-09-20 -10.628548  -9.153563 -7.883146  28.313940994  2002-09-21 -10.390377  -8.727491 -6.399645  30.914107995  2002-09-22  -8.985362  -8.485624 -4.669462  31.367740996  2002-09-23  -9.558560  -8.781216 -4.499815  30.518439997  2002-09-24  -9.902058  -9.340490 -4.386639  30.105593998  2002-09-25 -10.216020  -9.480682 -3.933802  29.758560999  2002-09-26 -11.856774 -10.671012 -3.216025  29.369368[1000 rows x 5 columns]
HDF5

Reading and writing to HDFStores

Writing to a HDF5 Store

In [138]: df.to_hdf('foo.h5','df')

Reading from a HDF5 Store

In [139]: pd.read_hdf('foo.h5','df')Out[139]:                    A          B         C          D2000-01-01   0.266457  -0.399641 -0.219582   1.1868602000-01-02  -1.170732  -0.345873  1.653061  -0.2829532000-01-03  -1.734933   0.530468  2.060811  -0.5155362000-01-04  -1.555121   1.452620  0.239859  -1.1568962000-01-05   0.578117   0.511371  0.103552  -2.4282022000-01-06   0.478344   0.449933 -0.741620  -1.9624092000-01-07   1.235339  -0.091757 -1.543861  -1.084753...               ...        ...       ...        ...2002-09-20 -10.628548  -9.153563 -7.883146  28.3139402002-09-21 -10.390377  -8.727491 -6.399645  30.9141072002-09-22  -8.985362  -8.485624 -4.669462  31.3677402002-09-23  -9.558560  -8.781216 -4.499815  30.5184392002-09-24  -9.902058  -9.340490 -4.386639  30.1055932002-09-25 -10.216020  -9.480682 -3.933802  29.7585602002-09-26 -11.856774 -10.671012 -3.216025  29.369368[1000 rows x 4 columns]
Excel

Reading and writing to MS Excel

Writing to an excel file

In [140]: df.to_excel('foo.xlsx',sheet_name='Sheet1')

Reading from an excel file

In [141]: pd.read_excel('foo.xlsx','Sheet1',index_col=None,na_values=['NA'])Out[141]:                    A          B         C          D2000-01-01   0.266457  -0.399641 -0.219582   1.1868602000-01-02  -1.170732  -0.345873  1.653061  -0.2829532000-01-03  -1.734933   0.530468  2.060811  -0.5155362000-01-04  -1.555121   1.452620  0.239859  -1.1568962000-01-05   0.578117   0.511371  0.103552  -2.4282022000-01-06   0.478344   0.449933 -0.741620  -1.9624092000-01-07   1.235339  -0.091757 -1.543861  -1.084753...               ...        ...       ...        ...2002-09-20 -10.628548  -9.153563 -7.883146  28.3139402002-09-21 -10.390377  -8.727491 -6.399645  30.9141072002-09-22  -8.985362  -8.485624 -4.669462  31.3677402002-09-23  -9.558560  -8.781216 -4.499815  30.5184392002-09-24  -9.902058  -9.340490 -4.386639  30.1055932002-09-25 -10.216020  -9.480682 -3.933802  29.7585602002-09-26 -11.856774 -10.671012 -3.216025  29.369368[1000 rows x 4 columns]
Gotchas

If you are trying an operation and you see an exception like:

>>> ifpd.Series([False,True,False]):    print("I was true")Traceback    ...ValueError: The truth value of an array is ambiguous. Use a.empty, a.any() or a.all().

See Comparisons for an explanation and what to do.

See Gotchas as well.




https://wap.sciencenet.cn/blog-1271266-858777.html

上一篇:层次聚类算法 各种linkage
下一篇:Python中的结构化数据分析利器-Pandas简介
收藏 IP: 14.204.63.*| 热度|

0

该博文允许注册用户评论 请点击登录 评论 (0 个评论)

数据加载中...

Archiver|手机版|科学网 ( 京ICP备07017567号-12 )

GMT+8, 2024-3-29 19:54

Powered by ScienceNet.cn

Copyright © 2007- 中国科学报社

返回顶部