I have a pandas dataframe(df) like this
Close Close Close Close Close Date 2000-01-03 00:00:00 NaN NaN NaN NaN -0.033944 2000-01-04 00:00:00 NaN NaN NaN NaN 0.0351366 2000-01-05 00:00:00 -0.033944 NaN NaN NaN -0.0172414 2000-01-06 00:00:00 0.0351366 -0.033944 NaN NaN -0.00438596 2000-01-07 00:00:00 -0.0172414 0.0351366 -0.033944 NaN 0.0396476 in R If I want to select fifth column
five=df[,5] and without 5th column
rest=df[,-5] How can I do similar operations with pandas dataframe
I tried this in pandas
five=df.ix[,5] but its giving this error
File "", line 1 df.ix[,5] ^ SyntaxError: invalid syntax 4 Answers
Use iloc. It is explicitly a position based indexer. ix can be both and will get confused if an index is integer based.
df.iloc[:, [4]] 
For all but the fifth column
slc = list(range(df.shape[1])) slc.remove(4) df.iloc[:, slc]
or equivalently
df.iloc[:, [i for i in range(df.shape[1]) if i != 4]] If you want the fifth column:
df.ix[:,4] Stick the colon in there to take all the rows for that column.
To exclude a fifth column you could try:
df.ix[:, (x for x in range(0, len(df.columns)) if x != 4)] If your DataFrame does not have column/row labels and you want to select some specific columns then you should use iloc method.
example if you want to select first column and all rows:
df = dataset.iloc[:,0] Here the df variable will contain the value stored in the first column of your dataframe.
Do remember that
type(df) -> pandas.core.series.Series Hope it helps
To select filter column by index:
In [19]: df Out[19]: Date Close Close.1 Close.2 Close.3 Close.4 0 2000-01-0300:00:00 NaN NaN NaN NaN -0.033944 1 2000-01-0400:00:00 NaN NaN NaN NaN 0.035137 2 2000-01-0500:00:00 -0.033944 NaN NaN NaN -0.017241 3 2000-01-0600:00:00 0.035137 -0.033944 NaN NaN -0.004386 4 2000-01-0700:00:00 -0.017241 0.035137 -0.033944 NaN 0.039648 In [20]: df.ix[:, 5] Out[20]: 0 -0.033944 1 0.035137 2 -0.017241 3 -0.004386 4 0.039648 Name: Close.4, dtype: float64 In [21]: df.icol(5) /usr/bin/ipython:1: FutureWarning: icol(i) is deprecated. Please use .iloc[:,i] #!/usr/bin/python2 Out[21]: 0 -0.033944 1 0.035137 2 -0.017241 3 -0.004386 4 0.039648 Name: Close.4, dtype: float64 In [22]: df.iloc[:, 5] Out[22]: 0 -0.033944 1 0.035137 2 -0.017241 3 -0.004386 4 0.039648 Name: Close.4, dtype: float64 To select all columns except index:
In [29]: df[[df.columns[i] for i in range(len(df.columns)) if i != 5]] Out[29]: Date Close Close.1 Close.2 Close.3 0 2000-01-0300:00:00 NaN NaN NaN NaN 1 2000-01-0400:00:00 NaN NaN NaN NaN 2 2000-01-0500:00:00 -0.033944 NaN NaN NaN 3 2000-01-0600:00:00 0.035137 -0.033944 NaN NaN 4 2000-01-0700:00:00 -0.017241 0.035137 -0.033944 NaN 
