Q: Not able to install DTW algorithm in DSX's R notebook - data-science-experience

Question

I am trying to install DTW package of R in DSX's R notebook.
when running install command:
install.packages("dtw")
gives following error:
"installation of package ‘dtw’ had non-zero exit status” warning.

Answer 1

Additional output from the installation attempt can be found in the log file for your R kernel. In the R notebook, execute
Sys.getenv("LOG_FILE")
to get the name of the log file. Then open a Python notebook and execute
!cat <name-of-the-log-file>
to see the contents of the log file.
After trying to install dtw, my log file showed:
gcc -std=gnu99 -shared -L/usr/local/lib64 -o dtw.so computeCM.o Launching java with spark-submit command /usr/local/src/spark20master/spark/bin/spark-submit sparkr-shell /gpfs/fs01/user/s10d-7f8820dde34dec-4cf890276e2b/notebook/tmp/RtmpofrZnw/backend_portd1c4793df5e
which looks like garbled log output from two processes writing to the same log file. But a bit further down, I find:
gcc: error: Launching: No such file or directory
gcc: error: java: No such file or directory
gcc: error: with: No such file or directory
gcc: error: spark-submit: No such file or directory
gcc: error: command: No such file or directory
gcc: error: sparkr-shell: No such file or directory
gcc: error: /gpfs/fs01/user/s10d-7f8820dde34dec-4cf890276e2b/notebook/tmp/Rtmpkh93V2/backend_porte8853270b27: No such file or directory
make: *** [dtw.so] Error 1
ERROR: compilation failed for package ‘dtw’
So the garbled line was actually executed. I'd say that the installation logic for the dtw package makes some assumptions about the environment which are not satisfied in DSX.
To dig further into the problem, you'll have to debug the dtw installer, or find somebody who can do that for you.