Reading in ActiGraph CSV data

[astegner34]

Hello!

I am having difficulty reading in .gt3x files collected on GT3X+ monitors and so though if I converted them to .csv that I might have more luck. Using the defaults I was testing with a simple script:

GGIR(datadir = "C:/GGIR_Analysis/ID#/file.csv",
output = "C:/GGIR_Analysis/ID#/Results",
dataFormat = "actigraph_csv",
extEpochData_timeformat = "%m/%d/%Y %H:%M:%S",
windowsizes = c(5, 900, 3600),
overwrite = TRUE,
do.enmo = TRUE,
verbose = TRUE)

It will run parts 1 and 2, but then I get the following error in part 3:

g.sib.det(M, IMP, I, twd = c(-12, 12), acc.metric = params_general[["acc.metric"]], : Argument acc.metric is set to ENMO but not found in GGIR part 1 output data

I assumed that ENMO was the default. What am I missing in order to produce the converted ENMO data? Thanks for your help.

Aaron

[vincentvanhees]

For most post 2017 gt3x files GGIR should work. Are you using a specific ActiGraph model?

Are you working with the exported raw data or with exported count data? Your code indicates that you are working with exported count data, which is aggregated and not raw data. This means that ENMO is not available and you would have to use a different sleep detection algorithm tailored to the information available in count data.

If you want to extract ENMO data make sure to export raw acceleration data to csv and not count data, and to set dataFormat = "raw". If I remember correctly, GGIR automatically recognises the raw Actigraph csv format.

[vincentvanhees]

Closing, because no response from @astegner34
Please make an effort to engage with someone who tries to help you. You are wasting my time.