Select Columns based on Row value Pandas

I have a similar df1 with thousand columns and thousands rows. I would like to do a random sampling based on a condition in cells in row 1 (date0) Basically i would like to random sample the columns and return them and the Datetime index based on the condition if the cell on date0 row == V1 and then do the same sampling for cell==V3 and then VX..... Then I would concatenate all those samples into a singular dataframe. I want to make sure I return the original Datetime Index and not a generic Index 0,1,2,3...

        abc def ghi jkl mno pqr
date0   V1  V1  V2  V3  V0  V1

date1  2  5    6   3   2    1

date2 3  1    1    3   5    6

date3 4  4    2    7   8   0

I have tried this so far but it does not work Dataset1= Dataset.ix[:,(random.sample(list(Dataset.iloc[0,:]=='V2'), SAMPLE_NUMBER))].copy()

SAMPLE_NUMBER is a random number for the number of columns to return.. thanks!