CODEWITHSUNDEEP
pythonpandas
user2989613
user2989613

Reputation: 1251

Pandas index value relating to a threshold exceedance

I am interested in reporting that date at which a threshold value is exceeded in multiple time series columns. The index is a date time value and column names related to site locations to which the time series data relate. I require a function similar to "idxmax" but to return index at which a threshold is FIRST exceeded in the time series. This seems an easy task but I am a new Python user and need a little guidance. Thanks.

Upvotes: 2

Views: 2125

Answers (1)

unutbu
unutbu

Reputation: 879561

It seems like idxmax could fit the bill:

In [44]: x = pd.Series([1,2,3,4,5], index=pd.date_range('2000-1-1', periods=5, freq='M'))

In [45]: x
Out[45]: 
2000-01-31    1
2000-02-29    2
2000-03-31    3
2000-04-30    4
2000-05-31    5
Freq: M, dtype: int64

In [46]: x >= 3
Out[46]: 
2000-01-31    False
2000-02-29    False
2000-03-31     True
2000-04-30     True
2000-05-31     True
Freq: M, dtype: bool

In [47]: (x >= 3).idxmax()
Out[47]: Timestamp('2000-03-31 00:00:00', tz=None)

Upvotes: 5

Related Questions