Downsampling of the event segment, while saving in the metadata.

[Pratyma]

Recently, when I tried to extract one of the events from the metadata (sqlite database), I found that the time series in the event was downsampled by almost a factor of 30. I have 15000 samples for the same event in my signal, but the extract from metadata has only 499 samples. The shape of the event's waveform is preserved. Both of them look exactly the same.

[abalijepalli]

I can't seem to reproduce this. Below are screenshots of two representative events. The blue trace is the raw data and the orange is from the database. They seem to overlap perfectly with no evidence of downsampling. Could you post similar screenshots of what you observe?

[Pratyma]

Now, I am also getting the same number of samples from metadata's time series, when I directly extracted. But previously, the events were extracted from mosaic_plot/mosaicAnalysis/analysisTimeSeries.py, which also uses the same sqlite database.

[abalijepalli]

I don't have that file. Is this in the repository? If so could you post a link to the code?

[Pratyma]

Apology for the wrong file name, it is mosaicweb/mosaicAnalysis/analysisTimeSeries.py.

I think I also found why decimation occurs. I think it is due to line 106-110 in mosaicweb/mosaicAnalysis/analysisTimeSeries.py.

def _calculateDecimation(self, dataLen): if dataLen < 1000: return 1 else: return int(round(dataLen/500.))

Therefore, it was not a bug. I missed the above-mentioned function. Sorry for the misunderstanding.

[abalijepalli]

That is only used when displaying long time series in the user interface. I'll close this issue.