[Python] How to calculate percentile rank of numpy array like in pandas

I have a Pandas DataFrame with 600 rows and 100 columns. I currently use this code to calculate rank in pandas:

df[rankCols].transform('rank', pct=True)


But I want to calculate it in numpy to improve speed. As I later convert my DataFrame to numpy array.

How to calculate the same in numpy? I can rank using argsort() but it does not calculate percentile rank.

Continue reading...