Python dataframe - count occurrences in a specified range (not axis!)

问题

I have a dataframe (called df), where there is a time series with a time stamp (first column) and several integer data columns.

TimeStamp Country 1 Country 2

12:00:00 10.05 21.60

11:59:00 11.12 22.33

11:58:00 12.18 21.70

11:57:00 11.70 21.60

11:56:00 11.65 22.33

11:55:00 11.70 21.60

11:54:00 11.50 22.33

11:53:00 11.80 21.80

... ... ...

Problem: I'd like to count the number of occurrences of a the maximum in a specific range (not the whole axis!).

E.g. In column Country 2, I'd like to count the number of occurrences of the max value in rows 1-8. So first I find the max value: df.iloc[0:7,1].max() -> 22.33

and now I'd like to count them, how to do it?

I'm looking for something like count(range, target value)

-> df.count(df.iloc[0:7,1)], df.iloc[0:7,1].max())

The output should be an integer. Here the max value (which is 22.33) occurs 3 times in the defined range, so I'd expect 3.

Thanks for your help

回答1:

Compare each value of filtered Series by maximum and count Trues values by sum:

s = df.iloc[0:7,1]

count = s.eq(s.max()).sum()
#alternative
count = (s == s.max()).sum()

print (count)
3

EDIT: Use Series.between:

s = df.iloc[0:7,1]

thr = 0.01
#print (s.max() - thr)
#print (s.max() + thr)

count = s.between(s.max() - thr, s.max() + thr).sum()
print (count)
3

来源：https://stackoverflow.com/questions/61287698/python-dataframe-count-occurrences-in-a-specified-range-not-axis