Skip to content

Latest commit

 

History

History
54 lines (49 loc) · 2.23 KB

Gurobi optimizer with R.md

File metadata and controls

54 lines (49 loc) · 2.23 KB
Raw

Gurobi optimizer with R

Gurobi optimizer is a problem solving software that can be used within R. It can solve integer, linear, and quadratic programming optimizations. These techniques can help to find the answers to complex models.

Example of running Gurobi optimizer with R at CARC

There are modules for both Gurobi and R on the wheeler cluster. All you need to do is load them, and then start an R session. This command is for version 8.1.0, however there are other versions of gurobi available (enter module avail gurobi to see a full list).

username@wheeler-sn:~$ module load gurobi/8.1.0
username@wheeler-sn:~$ module load r-3.6.0-gcc-7.3.0-python2-7akol5t
username@wheeler-sn:~$ R

Once you have started an R session, you can install packages just as you would in R. If you ever run into issues loading packages in R at CARC, you can reach out for assistance by emailling [email protected]. One piece of advice if you are using JupyterHub to run an R notebook at CARC is you may need to install packages from ther terminal window on JupyterHub because the notebook will not let you interactiively answer questions installs may need.

Start by installing the gurobi package:

> install.packages('/opt/local/gurobi/8.1.0/linux64/R/gurobi_8.1-0_R_3.5.0.tar.gz')
Installing package into '/users/username/R/x86_64-pc-linux-gnu-library/3.6'
* installing *binary* package 'gurobi' ...
* DONE (gurobi)

You should now be able to load the gurobi library in an R session:

> library(gurobi)
Loading required package: slam

Note that if you get an error regarding slam, you can install it using the command:

install.packages("slam", repos = "https://cloud.r-project.org")

Now let's runs a quick model as an example of what Gurobi can do and to see if everything is working properly:

> model <- list()
> model$A          <- matrix(c(1,2,3,1,1,0), nrow=2, ncol=3, byrow=T)
> model$obj        <- c(1,1,2)
> model$modelsense <- 'max'
> model$rhs        <- c(4,1)
> model$sense      <- c('<', '>')
> model$vtype      <- 'B'
> params <- list(OutputFlag=0)
> result <- gurobi(model, params)
> print('Solution:')
[1] "Solution:"
> print(result$objval)
[1] 3
> print(result$x)
[1] 1 0 1