I had successfully installed Microsoft Machine Learning Server in Ubuntu 16.04 with a one-box configuration. Everything worked till the point where i had to consume the deployed web service. The first error that showed was it couldn't find the package dplyr,
hence, i installed dplyr package in this path /opt/microsoft/mlserver/9.2.1/libraries/RServer as follows:
install.packages("dplyr",dependencies=TRUE,lib= "/opt/microsoft/mlserver/9.2.1/libraries/RServer")
Then the following error was thrown
[1] "Error: package or namespace load failed for ‘dplyr’ in loadNamespace(j <- i[[1L]], c(lib.loc, .libPaths()), versionCheck = vI[[j]]):\n there is no package called ‘R6’\npred could not be returned.\n"
I restarted the r session and tried to install R6 and the following Error shows
install.packages("R6",dependencies=TRUE,lib="/opt/microsoft/mlserver/9.2.1/libraries/RServer")
% Total % Received % Xferd Average Speed Time Time Time Current
Dload Upload Total Spent Left Speed
100 315k 100 315k 0 0 170k 0 0:00:01 0:00:01 --:--:-- 170k
No protocol specified
* installing *source* package ‘R6’ ...
** package ‘R6’ successfully unpacked and MD5 sums checked
** R
** inst
** preparing package for lazy loading
Error in unloadNamespace(pkg_name) :
namespace ‘R6’ is imported by ‘CompatibilityAPI’ so cannot be unloaded
* removing ‘/opt/microsoft/mlserver/9.2.1/libraries/RServer/R6’
The downloaded source packages are in
‘/tmp/Rtmpi43m3t/downloaded_packages’
Warning message:
In install.packages("R6", dependencies = TRUE, lib = "/opt/microsoft/mlserver/9.2.1/libraries/RServer") :
installation of package ‘R6’ had non-zero exit status
I restarted the R Session and tried to unload "CompatibilityAPI" from the Namespace (including the packages that "ComaptibilityAPI" depends on like RevoScaleR and MicrosoftML; also unloaded R6), tried to re-install the R6 again unfortunately
ends up throwing the same error and warning message.
What could be the solution to this?
