Minimum accelerometer recorded days for GGIR

[廖耿泽]

Hi all,

I have some GENEActiv only recorded for three days. Is it suitable to apply GGIR to these data?

Best regards,

Gengze

[Vincent van Hees]

Hi Gengze,

That depends. Technically GGIR can process recordings that last only half an hour if you are only interested in the time series. For GGIR part2 you need at least one valid calendar day, for sleep analysis (part3+4) you need at least one valid night, and for combined analysis you need at least 2 consecutive nights (part 5).

However, I think your question is whether 3 days is enough to answer your research question. This depends on your research question, the sample size, the population you are studying and their typical day-to-day variation in behaviour, and the statistical analysis you intend to use.

This is probably not a GGIR discussion. So, I think you are better off looking up how studies in your field motivated their participant inclusion criteria and/or contacting researchers.

Best regards,

Vincent

Dr. Vincent van Hees | Independent consultant | https://accelting.com/

------- Original Message -------

--
You received this message because you are subscribed to the Google Groups "R package GGIR" group.
To unsubscribe from this group and stop receiving emails from it, send an email to RpackageGGIR...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/RpackageGGIR/334af9fa-11b2-4456-b494-29ed906602d9n%40googlegroups.com.

[廖耿泽]

Dear Vincent,

Thank you very much for your detailed reply. Actually, we set 100 Hz for 7 days, but due to the shortage of the equipment, we had to let the participants return the GENEActiv on day 4. Therefore, we have 3 consecutive days of valid data. 

In this case, can I use the following arguments to "cut" the data into complete and valid 3 days?

              

               ..........

              #part2#
               hrs.del.start = 0,          hrs.del.end = 0,

               maxdur = 4,                 

               ..........

I am looking forward to your reply.

Best regards,

Gengze 

[Vincent van Hees]

Sounds correct, but you may want to check the pdf output in the results/QC folder produced after running GGIR part2, as that will clarify whether the data is correctly masked.

Vincent

Dr. Vincent van Hees | Independent consultant | https://accelting.com/

------- Original Message -------

To view this discussion on the web, visit https://groups.google.com/d/msgid/RpackageGGIR/ed159e74-00f2-4e17-b01c-c4cda37021f0n%40googlegroups.com.

[廖耿泽]

Dear Vincent,

I noticed another issue with the sleep data. In the seven-day data, I found that most “part4_nightsummary_sleep_cleaned” miss the first- or last-day of sleep data.  Almost all the files missed the first-day data in the three-day data.

Please help me check the code.

 
Here is the code:
g.shell.GGIR(mode=c(1,2,3,4,5),
                      datadir=….,
                      outputdir=….,
                      studyname = ….,
                      do.report=c(2,4,5),
                     
                      #part2#
                      strategy = 1,

                      hrs.del.start = 0,          hrs.del.end = 0,

                      maxdur = 9,  

#maxdur = 9 for seven-day data; maxdur = 3 for three-day data

                     includedaycrit = 16,
                      qwindow=c(0,24),
                      mvpathreshold =c(100),
                      bout.metric = 4,
                      excludefirstlast = FALSE,
                      includenightcrit = 16,
                     
                      #part3 and 4#
                      def.noc.sleep = 1,
                      outliers.only = TRUE,
                      criterror = 4,
                      do.visual = TRUE,
                      ignorenonwear=T,
                     
                      #part 5#
                      threshold.lig = c(30), threshold.mod = c(100),  threshold.vig = c(400),
                      boutcriter = 0.8,      boutcriter.in = 0.9,     boutcriter.lig = 0.8,
                      boutcriter.mvpa = 0.8, boutdur.in = c(1,10,30), boutdur.lig = c(1,10),
                      boutdur.mvpa = c(1),
                      includedaycrit.part5 = 2/3,
                     
                      #visual report#
                      timewindow = c("WW"),
                      visualreport=TRUE)

[Vincent van Hees]

Dear Gengze,

Your code looks good. By specifying argument includenightcrit = 16​ you ask GGIR to only estimate sleep from nights with at least 16 hours of data between noon and the following noon (if the person wakes up after noon then the window is defined between 6pm and 6pm).

Your pdf visualisations indicates your recordings started after 6pm, which makes it impossible for GGIR to find 16 or more valid hours before the next noon.

One solution is that you lower the value for includenightcrit​ to for example 14, although that could compromise the reliability of estimates for other recording.

A better but less quick solution is that we develop an enhancement of GGIR to consider a late start of the recording when trying to detect sleep for the first night. Here, I see the following options:

• You or someone else develops this as a contribution to GGIR.
• You or someone else hires me to develop this.
• I keep put it on my list of ideas for the future and maybe at some point in the upcoming year(s) I can do the work as part of another project.

Best regards,

Vincent

Dr. Vincent van Hees | Independent consultant | https://accelting.com/

------- Original Message -------

To view this discussion on the web, visit https://groups.google.com/d/msgid/RpackageGGIR/55b1a477-73f9-44a6-a7f6-ad2a5fe8c440n%40googlegroups.com.

[Vincent van Hees]

Dear Shashi,

Please post your questions in the google group (cc). I only provide free support in public as that enables others learn from my answers, it reduces the chance that I have to answer the same question twice, and it gives recognition for the time I put into this.

To get only the part 2 report you will need to have processed the data with part 1 and 2 based on argument mode = c(1, 2) or mode = 1:2 which is the same.

Next, to only create part 2 report use do.report = 2

If you already have part 2 milestone data, use arguments: mode = c() and do.report = 2.

Vincent

Dr. Vincent van Hees | Independent consultant | https://accelting.com/

------- Original Message -------
On Tuesday, December 20th, 2022 at 6:39 PM, shashi dhar <sd5...@gmail.com> wrote:

Hi Vincent,

How to run only part 2 report?

To view this discussion on the web, visit https://groups.google.com/d/msgid/RpackageGGIR/2HeJH1m197K6t2d-q_Mfwizgo2FXGKKmOT0F9f-Ge8eLKFlBZB7XLd2SDa8OGkDavIDAJvTKH1d3A5ycQfL8Joa0rawXMlDMy9EdfdJl0g4%3D%40accelting.com.