Error report while running

[Jinyu8579]

Hi!

I am not sure if this is a right place to attach the problem here. I filled all the parameters in diyabcGUI and when I run it, it returns an error (see attach txt). Dose this mean that the scenario I construct can not be reasonably verified?

error report.txt

Thanks, Jinyu

[gdurif]

Hi, You posted at the right place. Sorry for the late response, I am under quite a rush for the new year. I'll look into it and get back to you ASAP. Best,

[gdurif]

Sorry for the late response. Could you tell me more about the type of data you were considering ? (MSS, SNP indseq/poolseq)

If MSS data, I think there is a potential link with #124 and https://github.com/diyabc/diyabc/issues/61

[gdurif]

FYI, we just published a new release (version 1.2.1) which should fix several issues. Could you tell me if it works better? Thanks in advance, Best regards

[Jinyu8579]

Hi, we encounter the problem in the analyses of MSS data. I suspect the problem may be related with the conditional constraints on merge/split time. Because the DIYABC-RF program (shiny application or web service) will disappear when we try to write conditional constraints (e.g., t2>t1, we copy the ">" from panel in the "Condition setting" frame to avoid mistakes) in the "Condition setting" frame, we have to let the "Condition setting" frame be blank (i.e., let the scenarios constructed without conditional constraints on merge/split time). Then, we find that the DIYABC-RF program will work for the analyses only considering some simple scenarios (e.g., only merge), while it fails to work when the analyses adding some other complicated scenarios (e.g., split). Thus, finally, we settle the problem using previous DIYABC 2.1.0 to generate the project files and applying DIYABC-RF program to run the project (work well). FYI, we cannot successfully launch DIYABC-RF_GUI by running DIYABC-RF_GUI (or DIYABC-RF_GUI.bat). When we click on DIYABC-RF_GUI (or DIYABC-RF_GUI.bat), the GUI will appear and then disappear immediately. I suspect this may be related with the Chinese characters exist in the computer we used (we are in China). We then launch DIYABC-RF in R and encounter the previous reported problems. Then problems appear when we use DIYABC-RF either as a shiny application or hosted as a web service (we have tried Chorme, Microsoft Edge, and Internet Explorer).

[Jinyu8579]

BTW, we settle the problem using previous DIYABC 2.1.0 to generate the project files and applying DIYABC-RF program to run the project. We conduct the analyses by using microsatellite loci and DNA sequence loci, respectively. It works well for most of the analyses (Scenario choice, parameter estimation), except that it reports errors when we try Prior-scenario checking PRE-ANALYSIS for DNA sequence loci.

[Jinyu8579]

Sorry for delayed reply. The resubmission manuscript is on the deadline.

[gdurif]

No problem, thanks for your detailed answer. I am glad you managed to find a solution to make it work. And again, sorry for the trouble.

We fixed some issues with the condition setting, and (numerous) other things... Hopefully the new release will be better. Sadly, the problem with the prior checking remains (c.f. https://github.com/diyabc/diyabc/issues/59) and also a more general problem with MSS training set data simulation (c.f. https://github.com/diyabc/diyabc/issues/61). These two points are not specific to the GUI though.

Thanks again

[dididididisco]

Hi, we encounter the problem in the analyses of MSS data. I suspect the problem may be related with the conditional constraints on merge/split time. Because the DIYABC-RF program (shiny application or web service) will disappear when we try to write conditional constraints (e.g., t2>t1, we copy the ">" from panel in the "Condition setting" frame to avoid mistakes) in the "Condition setting" frame, we have to let the "Condition setting" frame be blank (i.e., let the scenarios constructed without conditional constraints on merge/split time). Then, we find that the DIYABC-RF program will work for the analyses only considering some simple scenarios (e.g., only merge), while it fails to work when the analyses adding some other complicated scenarios (e.g., split). Thus, finally, we settle the problem using previous DIYABC 2.1.0 to generate the project files and applying DIYABC-RF program to run the project (work well). FYI, we cannot successfully launch DIYABC-RF_GUI by running DIYABC-RF_GUI (or DIYABC-RF_GUI.bat). When we click on DIYABC-RF_GUI (or DIYABC-RF_GUI.bat), the GUI will appear and then disappear immediately. I suspect this may be related with the Chinese characters exist in the computer we used (we are in China). We then launch DIYABC-RF in R and encounter the previous reported problems. Then problems appear when we use DIYABC-RF either as a shiny application or hosted as a web service (we have tried Chorme, Microsoft Edge, and Internet Explorer).

我靠 我也是，点一下就闪运行然后就无了，我找了个实验室win10的电脑运行就没遇见过这个问题。