Inf in data

[roaldarbol]

Inf in data

[mooch443]

Hey!
Inf simply stands for "no data" and can be understood synonymously to NaN, so your assumption was correct. It is indeed a bit worrying that there seem to be outlier values after. Your best bet right now will be to remove them, yes, but I will investigate this. Thanks for pointing out!
-Tristan

[mooch443]

The next release (v1.1.2) will add a new option output_invalid_value which can be set to NaN. This way, you'll be able to achieve the desired behaviour. I also fixed the issue that you pointed at with the large outlier values. This is fixed now - as intended previously, this uses the (correct) time since the individual was lost to when it was found again, to interpolate speed. You may be right that it is advisable to remove this first value in your analysis (if this is critical), but at least it won't add outliers anymore (at least according to my testing).

This issue will close with the release of the new version, thanks again for reporting this!