search

smousavi05 / EQTransformer

EQTransformer, a python package for earthquake signal detection and phase picking using AI.
https://rebrand.ly/EQT-documentations
MIT License
305 stars 148 forks source link

Using different filter range than the training dataset #92

Closed hasbishiddiqi closed 2 years ago

hasbishiddiqi commented 2 years ago

Hi @smousavi05,

I notice that in mseed_predictor, a band pass filter is apllied to the waveforms (1 - 45 Hz). I guess this is because the training dataset was also filtered in this frequency range.

I have some noisy stations with high frequency noise. The signal become clearer if I filter it with lower frequency raange (for example: max frequency 1-10 Hz or 1-15 Hz). So I modified the filter range in the mseed_predictor, and I got more picks from these noisy stations. I use the EqT_model.h5 training model.

Do you think this is OK or this can cause some issues?

Thanks!

Hasbi

smousavi05 commented 2 years ago

@hasbishiddiqi Hi Hasbi, that is perfectly fine as you have just narrowed the frequency band. This is totally okay and should not make any problem.

hasbishiddiqi commented 2 years ago

@smousavi05 Thank you!