import pandas as pd
import numpy as np
import math
[docs]def Rr(rank):
if(np.isnan(rank)):
return 0
else:
return 1.0/rank
[docs]def RrScore(rankings):
"""
Reciprocial rank, see https://en.wikipedia.org/wiki/Mean_reciprocal_rank .
"""
return rankings['rank'].map(Rr)