[FieldTrip] indepsamplesF produces NaN

[Schoffelen, J.M. (Jan Mathijs)]

Dear Enya,

Assuming that all went well with the definition of the design matrix etc, I suspect that there are NaNs in your data that cause you troubles.

The difference between the ‘ft_statfun_indepsamplesT’, and ‘ft_statfun_indepsamplesF’ is that the former knows how to handle NaNs in the data, whereas the latter doesn’t. My first hunch would therefore be that your data contains NaNs or Infs. Even a single NaN (or inf) would spoil it for the rest of the well-behaved data samples.

Best wishes,
Jan-Mathijs

On 16 Mar 2021, at 15:07, Enya Weidner <enya.weidner at uni-bielefeld.de<mailto:enya.weidner at uni-bielefeld.de>> wrote:

Hi everyone!

I am currently trying to run statistical analyses on timelocked data by comparing the values of three different experimental conditions with help of an F-Test. I tried using the “indepsamplesF” function and it runs smoothly with no errors (there is a warning when I run the montecarlo method saying: “no clusters were found in the observed data”) but produces NaNs in the probabilities and test-statistics. This happens for both the “montecarlo” and the “analytic” method. If I run just simple t-Tests I am getting interpretable results. Does this simply indicate that there are no significant differences between conditions? Wouldn’t there still be an F-value indicating such a result?

Any ideas as to why this happens?

Thanks a lot in advance for your help!

Enya Weidner
Universität Bielefeld
Fakultät für Psychologie und Sportwissenschaft
AE02 Affektive Neuropsychologie



----------------------
M.Sc. Enya Weidner
Universität Bielefeld
Fakultät für Psychologie und Sportwissenschaft
AE02 Affektive Neuropsychologie

Raum: T5-220
Telefon: 0521-106 4533
Email: enya.weidner at uni-bielefeld.de<mailto:enya.weidner at uni-bielefeld.de>


_______________________________________________
fieldtrip mailing list
https://mailman.science.ru.nl/mailman/listinfo/fieldtrip
https://doi.org/10.1371/journal.pcbi.1002202

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20210317/621fc604/attachment.htm>