search

numenta / NAB

The Numenta Anomaly Benchmark
GNU Affero General Public License v3.0
1.93k stars 869 forks source link

How to consider probationary data #381

Open ir1979 opened 3 years ago

ir1979 commented 3 years ago

Hi, how to reproduce the results of the "null" detector such that "total_records_in_dataset - sum(TP, TN, FP, FN) = 750", i.e., the scorer skips the probationary part of the input data? Is there any specific switch (line argument) for run.py? Does the detector have to handle the first 750 records, itself?

markNZed commented 3 years ago

See https://github.com/numenta/NAB/blob/2fafea9c37e250edf5281d7f23627dc5e4ebfcac/nab/sweeper.py#L194