Data output in tabular form with timestamps and derived "sleep/rest" versus "wake/active" binary-data column?

[jthebl]

Hey everyone!

I was wondering if there is an output within the numerous GGIR outputs (or relevant intermittent steps between the outputs) where it is possible to extract a tabular dataframe (e.g. a .csv) that would be on a per participant/subject basis that has timestamps as the observations/rows, and in a separate column the rest/wake-algorithm-derived (e.g. the GGIR default algorithm for SIB detection: vanHees2015) outputs that correspond to those timestamps? I would think it would be the same data used to generate the "file summary reports" but instead of an actogram the data could be accessed via a tabular dataframe.

For sleep/wake detection I was previously using the acigraph.sleepr package with apply_choi() and apply_cole_kripke() for non-wear detection and subsequent Sleep/Wake identification, respectively. However, I would like to migrate more of my methods to the GGIR package. My current project/analysis involves data collected from firefighters in which I hope to use GGIR-based rest/activity detection algorithm(s), in addition to data quality/non-wear detection methods, in conjunction with HR data collected from a separate device (polar), and also recorded "alarm/calls" that the particular firefighter-subject responded to during the data-collection period.

Thanks for the help!
Joey

[vincentvanhees]

Yes, GGIR exports time series objects. See documentation for details: https://wadpac.github.io/GGIR/articles/chapter12_TimeUseAnalysis.html

I assume you already reached that conclusion and closed the discussion. However, for others with the same question it is still helpful to have the answer posted here.