smangul1/rop

Conda environment file with necessary packages

jaime11 opened this issue · 5 comments

jaime11 commented

Hello,

I am trying to install rop, thank you a lot for the software!
The issue is that I have conda installed already and the installer produces some errors due to pip and also tries to install miniconda. The server where I am trying to install the software doesn't have permission to write on certain areas (I can only write in my home directory). I am also reading from
https://github.com/smangul1/rop/wiki/How-to-install-ROP%3F

That the tool requires up to 50GB for some databases, so it would be necessary for my situation to be able to store this somewhere else since I don't have that amount of space in the drive.

Is it possible to export a conda environment file with the necessary packages to install to create a personalized environment to run the tool?

Thank you in advance

smangul1 commented

Kevin, Can you please help here

Serghei

kahsieh commented

@jaime11 The installer tries to install MiniConda by default, but the installation is entirely within rop's directory, so you shouldn't run into permission errors with MiniConda. That said, it can be disabled using the -n|--native option. The installer also offers the -d|--dbdest option to change the location where the database is stored, so you can select a location with more space.

I have prepared an environment.yml file. This should have everything needed for the conda environment.

jaime11 commented

Thank you, I realized about that raised issue afterwards. Just to note, the yml file in "tools" gave a pip install error for the package suffix_tree-2.1.tar.gz:

IOError: [Errno 2] No such file or directory

Since the installation installs this package later, I removed the install from the environment.yml

jaime11 commented

Note, as the environment.yml is now I am missing the package biopython

ImportError: No module named Bio

kahsieh commented

Thanks for the tips! The IOError probably occurred because the command to create the environment was invoked from a directory other than the one where environment.yml was located (the tools directory). To help avoid this, I moved the environment.yml to the root directory and updated it accordingly. I also added biopython to it.