If a column has 'x' value, find max value from other columns in row, and change x to the v-CodePudding

I have a dataframe similar to the following:

data = {'ref': ['0001', '0002', '0003', '0004', '0005', '0006', '0007'],
        'topic': [-1, 0, 1, 2, 0, -1, 2],
        '-1': [1, 0, 0, 0.1, 0, 0.99, 0],
        '0': [0, 1, 0, 0, 1, 0, 0.1],
        '1': [0, 0, 0.99, 0, 0, 0, 0],
        '2': [0.4, 0, 0, 0.7, 0, 0, 1],
        }

df = pd.DataFrame(data)

If the column topic value is -1, then I want to look in the same row of columns 0 to 2, and change the value in topic, to the header of the max value.

As an example, in the first row in the table above, the column topic has a value of -1, so I want to change that value to whatever is the header name of the max value in the row. So that would be column 2, which has a value of 0.4.

So now the table looks like this, where the topic value of ref 0001 has changed from -1 to 2

The other point is, that if like ref 0006, there is no value > 0 in the other columns despite having a topic value of -1, then it should be left alone.

I hope this makes sense. Struggling hard to get this done.

Thank you!

CodePudding user response：

def correct_topic(row):
    max_value = row[['0','1','2']].max()
    if max_value > 0:
        return row[row == max_value].index[0]
    else:
        return row.topic
    
df['topic'] = df.apply(correct_topic, axis=1)

I think this is what you need :)

CodePudding user response：

Setting up example:

import pandas as pd
data = {'ref': ['0001', '0002', '0003', '0004', '0005', '0006', '0007'],
        'topic': [-1, 0, 1, 2, 0, -1, 2],
        '-1': [1, 0, 0, 0.1, 0, 0.99, 0],
        '0': [0, 1, 0, 0, 1, 0, 0.1],
        '1': [0, 0, 0.99, 0, 0, 0, 0],
        '2': [0.4, 0, 0, 0.7, 0, 0, 1],
        }

df = pd.DataFrame(data)

Making a boolean mask to set the rows:

cols = [str(x) for x in range(0, 3)]
condition = (df.topic == -1) & ((df[cols] > 0).any(axis=1))

0     True
1    False
2    False
3    False
4    False
5    False
6    False
dtype: bool

using idxmax gets the names of the columns at the maximum values

df[cols].idxmax(axis=1)

0    2
1    0
2    1
3    2
4    0
5    0
6    2
dtype: object

Now you can use df.where and the boolean condition to set the topic values.

df["topic"] = df["topic"].where(~condition, df[cols].idxmax(axis=1)[condition])

Output:

    ref topic    -1    0     1    2
0  0001     2  1.00  0.0  0.00  0.4
1  0002     0  0.00  1.0  0.00  0.0
2  0003     1  0.00  0.0  0.99  0.0
3  0004     2  0.10  0.0  0.00  0.7
4  0005     0  0.00  1.0  0.00  0.0
5  0006    -1  0.99  0.0  0.00  0.0
6  0007     2  0.00  0.1  0.00  1.0