Different results when sda-filter is run within fit_ssm() vs. run independently?

[vheim]

Hello,

I am currently looking into fitting a SSM to my Argos data from SPOT-tag data. From reading the aniMotum vignette, I understood that the fit_ssm() function uses the trip::sda() filter to filter out spurious detections if the argument spdf = T. However, I realised that when I run the sda filter with spdf = T as part of the fit_ssm filter, the function removes a lot more detections compared to when I run trip::sda() with the same settings outside of fit_ssm() as an independent function call.

Is this something to be expected, i.e. do I interpret the inclusion of the sda filter within the fit_ssm() wrongly?

Best wishes,
Vital