I’m processing MRS data from two groups (control Vs Disease) that i want to compare. I actually can only process these groups separately and i’m looking to specify these two data groups at once in order to make better comparaisons (raincloud plots, correlation, etc…).
I read in the Ospery manual that this should be done via a csv file that we can specifiy in the jobfile, but i can’t find at wich section of the jobfile. Also, is there any csv file example that can help with the way that we should organize the csv file in order to include group category and external variables (age, behavioral data, etc…)?
Any help will be greatly appreciated.
Thanks for raising this issue. This functionality is not really documented well and I’ll work on that for the next release. Here’s how you can use the stat csv file:
Supply location of a csv file, which contains possible correlation measures and group variables. Each column must start with the name of the measure. For the grouping variable use ‘group’ and numbers between 1 and the number of included groups. If no group is supplied the data will be treated as one group.
Add the path to your job file like this:
file_stat = '.../statexample.csv';
with … being the path to your csv file. You can place this anywhere within the jobFile definition.
I’ve also attached an example file for you. The group name column is just for me to help identify each group.
Let me know whether this has solved your issues or if you have any other questions!
PS: I’d also try to keep the external variable names relatively simple. It is compiled through a Tex interpreter during the plot, so you could either try using Tex conform names or avoid ‘_’ and ‘^’ signs.
statexample.csv (67 Bytes)
Thank you for your quick reply and help.
I followed your instructions, Osprey can read the specified “stat.csv” file (please see attached csv file) but when i chekch with the overview GUI, the raincloud plots are not correctly plotted since plotted data points from two groups are affected to one group. Also, two new variables are created (“group” and “MRSCont.overview.corr”) and are non recognized when i try to plot them (please see “Overview” attached pdf file)
Another concern is about the mean spectra which is calculated for for all subjetcs as one group instead of separating them (please see the attached “AGrand_mean” pdf file).
Could you please help me with these issues?
Thank youAGrand_mean.pdf (56.5 KB) Overview.pdf (100.0 KB) stat.csv (59 Bytes)
Great! The csv file looks right and as you have mentioned the variables are added to the container.
There is a checkbox in the top right corner to turn on Grand mean or group separated plots (for the individual spectra, the mean spectra, or the raincloud plots). I have just tested this functionality and it should work fine.
Indeed there was a problem with the correlation with external variables. I have fixed this in the current version, just uploaded it into the develop branch.
Let me know if this has fixed your issues.