For these four tabs I have not created new server functions, instead I have modified the old one to include {bnma}. For consistency, this could also be done for the previous tabs in a later task.
I have not added the fancy numbering system, that will be the next task, as this is already quite a lot of change.
As well as removing the UME graph for covariate regression (and baseline risk) I have removed the UME model details in tabs 4b8 and 4c8.
I could not get the Gelman plots to display in a grid using bnma::network.gelman.plot, so I have recreated them manually and used par(mfrow = c(x,y)). Related to this, there were some occasions when not all the plots were displaying in the covariate regression. So when I create the grid size in bnma, I have also applied this to the covariate regression, which means all graphs are displayed in both cases.
To help with your review I have created a separate file to test a couple of things. One is that the Gelman plots mentioned in the previous point are faithfully recreated. The other is that the deviance statitistics in tab 4b7 are assigned to the correct studies. I thought this last check was necessary because I have reused the plots that took gemtc output as input, with bnma output passed to the functions as if they had come from gemtc.
The file is on Jira.
For these four tabs I have not created new server functions, instead I have modified the old one to include {bnma}. For consistency, this could also be done for the previous tabs in a later task.
I have not added the fancy numbering system, that will be the next task, as this is already quite a lot of change.
As well as removing the UME graph for covariate regression (and baseline risk) I have removed the UME model details in tabs 4b8 and 4c8.
I could not get the Gelman plots to display in a grid using bnma::network.gelman.plot, so I have recreated them manually and used par(mfrow = c(x,y)). Related to this, there were some occasions when not all the plots were displaying in the covariate regression. So when I create the grid size in bnma, I have also applied this to the covariate regression, which means all graphs are displayed in both cases.
To help with your review I have created a separate file to test a couple of things. One is that the Gelman plots mentioned in the previous point are faithfully recreated. The other is that the deviance statitistics in tab 4b7 are assigned to the correct studies. I thought this last check was necessary because I have reused the plots that took gemtc output as input, with bnma output passed to the functions as if they had come from gemtc. The file is on Jira.