Reputation: 1257
Is it possible to generate time based events that are unevenly distributed using Pandas? I'm currently using this, but am trying to have variation in how evenly these events are distributed across the series.
rng = date_range('1/1/2011', periods=60*3, freq='S')
ts = Series(0, index=rng)
Upvotes: 1
Views: 87
Reputation: 30424
Here's some example code based on the discussion in the comments. You can set '0.50' to be any percentage to keep.
import numpy as np
ts = Series( np.random.uniform(0,1,180), index=rng)
ts[ ts < 0.50 ].dropna()
2011-01-01 00:00:00 0.283262
2011-01-01 00:00:01 0.200472
2011-01-01 00:00:04 0.077024
2011-01-01 00:00:06 0.041452
Upvotes: 2