Change a pandas DataFrame column value based on another column value

Question

I have a dataframe with two columns each of which represents an organism. They are called ORG1 and ORG2 I want to move the values of ORG2 into ORG1 for the corresponding index value.

So, if ORG1 is 'A' and ORG2 is 'B' I want ORG1 to take the value 'B' from ORG2.

I have already started work to identify indexes of the ORG2 organisms that I want to move, as follows:

def move_org2(x):
    org2_matches = Series(x.ORG2.str.count("ESBL"))
    return x.ix[org2_matches == 1]

org2_DF = move_org2(DF)

org2_DF.ORG2.index

What is the best way to use this to change ORG1 values with the values at corresponding ORG2 indices

Answer 1

In [13]: df
Out[13]:
  ORG1  ORG2
0    A  ESBL
1    B     P
2    C     Q
3    D     R
4    E  ESBL

In [14]: cond = df.ORG2 == 'ESBL'

In [15]: df.ORG1[cond] = df.ORG2[cond]

In [16]: df
Out[16]:
   ORG1  ORG2
0  ESBL  ESBL
1     B     P
2     C     Q
3     D     R
4  ESBL  ESBL

Answer 2

In other words, using .loc you would do

In [2008]: df
Out[2008]:
  ORG1  ORG2
0    A  ESBL
1    B     P
2    C     Q
3    D     R
4    E  ESBL

In [2009]: df.loc[df['ORG2'] == 'ESBL', 'ORG1'] = df['ORG2']

In [2010]: df
Out[2010]:
   ORG1  ORG2
0  ESBL  ESBL
1     B     P
2     C     Q
3     D     R
4  ESBL  ESBL

Or, if you need a copy, without modifying original df , you can use .mask()

In [2016]: df.mask(df['ORG2'] == 'ESBL', df['ORG2'], axis=0)
Out[2016]:
   ORG1  ORG2
0  ESBL  ESBL
1     B     P
2     C     Q
3     D     R
4  ESBL  ESBL