R package (httk) cannot be installed during deployment

[PARODBE]

Hi,

I have problems installing the R httk package, the error says the following:

Warning in install.packages(“httk”, repos = “http://cran.us.r-project.org”, :[2024-03-09 11:48:33.265139]

[2024-03-09 11:48:33.265586] ‘lib = “/usr/lib/R/site-library”’ is not writable

Error in install.packages(“httk”, repos = “http://cran.us.r-project.org”, :

unable to install packages

Execution halted

I have followed the instructions of @dataprofessor in his video: https://www.youtube.com/watch?v=xTfaJsn2PN0&t=6s

And, another person (@123_Gao) had the same problem that mine:

R package (mirt) cannot be installed during deployment

Hi, this issue has been bothering me for days and I have tried some methods and still no solution. I failed to use r-cran-mirt in packages.txt and then checked Ubuntu/Debian to find that the package did not seem to be distributed. I tried subprocess.run([“Rscript”, “install.R”], check=True) which also failed and I was prompted with an unwritable message when I deployed it. Could I ask what else can be done to solve this issue? For example uploading a .tgz package to GitHub to install f…

However, he didn’t provide code or something in order to know how he solved the problem.

The link to my github is: GitHub - PARODBE/streamlit_col_new

Also, another question, is that in my computer (locally) I have written the following to import rscript via subprocess:

pro=subprocess.Popen([‘C:\Program Files\R\R-4.3.1\bin\x64\Rscript.exe’, “httk.R”], text=True, shell=True)

pro.wait()

However, this works locally but not in streamlit cloud if I put some print() as a test, in that case, in the cloud, I have to put shell=False like the code of @dataprofessor:

pro=subprocess.Popen([‘C:\Program Files\R\R-4.3.1\bin\x64\Rscript.exe’, “httk.R”], stdout=subprocess.PIPE, stderr=subprocess.PIPE, text=True,shell=False)

pro.wait()

Please could you help me?

[123_Gao]

I had problems integrating R with Python before. My final solution was to use Docker, which is more flexible. Deploy Streamlit using Docker - Streamlit Docs

[PARODBE]

Thank you for your response @123_Gao, but did you try of modifying the lib such as @Goyo suggested you? Or did you implement It directly with docker? Do you have GitHub for seeing your code? Thanks

[PARODBE]

One question @123_Gao , if I build my app via docker, do I need to do something different with the R packages and the code? Do I need the packages.txt file, or how did you do It? And did you run your code via subprocess or via rpy2?