Python Pandas: How do I average ONLY the data >1000 from several columns?

Try via filter(),gt() and mean():

out=df.filter(like='Revenue')
df['Average']=out[out.gt(1000)].mean(axis=1)

output of df:

    Name    Revenue1    Revenue2    Revenue3    Average
0   Peter   1000        2000        3000        2500.0
1   Jane    9000        10000       5000        8000.0

Breakdown of code:

Firstly we select all the columns that Named like 'Revenue' via filter() method and it will give a DataFrame of columns named like that so we are storing it in out variable

out=df.filter(like='Revenue')
#output of above code:
  Revenue1  Revenue2    Revenue3
0   1000    2000        3000
1   9000    10000       5000

Then we will filter out rows where values is greater than 1000:

out.gt(1000)
#your condition
#output of above code:
   Revenue1     Revenue2    Revenue3
0   False       True        True
1   True        True        True

As you saw the condition gives you boolean values so now we will pass that boolean mask to out so where there is True in the above boolean values you will get the values and where there is False you will get NaN this is know as boolean masking:

out[out.gt(1000)]
#output of above code:
   Revenue1     Revenue2    Revenue3
0   NaN         2000        3000
1   9000.0      10000       5000

Finally we will calculate mean via mean() method on axis=1 so the NaN's are ignored..In other words mean() method doesn't considered NaN's while calculating mean(because by default skipna=None in mean() method)

out[out.gt(1000)].mean(axis=1)
#output of above code:
0    2500.0
1    8000.0
dtype: float64

Finally we will assign this result back to df:

df['Average']=out[out.gt(1000)].mean(axis=1)

Update:

If df also has other numerical columns like 'Income' and you also want to include it in your calculation along with columns like 'Revenue' then use:

out=df.filter(regex='Revenue|Income')
df['Average']=out[out.gt(1000)].mean(axis=1)

or using a mask

df['average'] = df.mask(df.filter(like='Revenue')<=1000).mean(1)

   Name  Revenue1  Revenue2  Revenue3  Revenue4  average
0  Peter      1000      2000      3000      2500   2500.0
1   Jane      9000     10000      5000      8000   8000.0

Hello Try this out

 d['average']=d[d.filter(like='revenue')>1000].mean(1)
 d

Maybe try using a condition of > 1000:

df['Mean'] = df.select_dtypes(int)[df.select_dtypes(int) > 1000].mean(1)
print(df)

Output:

    Name    Revenue1    Revenue2    Revenue3    Mean
0   Peter   1000        2000        3000      2500.0
1   Jane    9000        10000       5000      8000.0