Quick tips

Expand

To make an environment from a file:
conda env create -f "RNAseq.yaml"

To make an enviornment:
conda create --name "name-of-env"

To list all conda environments:
conda env list

To move to a different conda environment:
conda activate "name-of-env"

To remove a conda environment:
conda remove --name myenv --all

To give jupyter a conda env install set:
python -m ipykernel install --user --name tensorflow --display-name "Python 3.8 (tensorflow)"

To update a conda env with a newer yml
activate your env
conda env update -f environment.yml

To export your current env to yml file
conda env export > environment.yml If you want a more succinct version (note doesn't keep channels?)
conda env export --from-history > environment.yml

If you want to use mamba from one env to create another env mamba install -n ops -y -c conda-forge mamba

Jupyterlab enhancements

Line-by-line

To allow for nicer running of R-code in JupyterLab https://stackoverflow.com/questions/56460834/how-to-run-a-single-line-or-selected-code-in-a-jupyter-notebook-or-jupyterlab-ce This seems like a healful way to create an Rstudio like python kernal.

Settings --> Advanced Settings Editor --> JSON Settings Editor then copy and paste this:

{
            "command": "notebook:run-in-console",
            "keys": [
                "Ctrl Shift Enter"
            ],
            "selector": ".jp-Notebook.jp-mod-editMode"
},

Connect to R

To give jupyter an R conda env follow potentially to add R kernels: https://stackoverflow.com/questions/68939097/how-to-use-different-versions-of-r-kernels-in-vs-code-jupyter-notebooks-when-usi

1. Make a conda env and get r-base
2. activate the environment
3. CD into ~/.local/share/jupyter/kernels and make a new directory with the same name as your conda env
4. Create a file called kernel.json

{"argv": 
        ["/SRA_store/shared/tools/mkozubov/miniconda3/envs/pcst/bin/R",
         "--slave",
         "-e",
         "IRkernel::main()",
         "--args",
         "{connection_file}"],
 "display_name":"Cytotalk-R 4.2.0",
 "language":"R"
}

fill the file with this, and make the R path the path to a specific conda R you want, and change the Cytotalk display name.

5. Make sure that the conda env, PCST in my case, has the irkernel conda installed otherwise the kernel just wont connect!
6. If we already have an R installed on our device, we can do install.packages('IRkernel') in it, then pass the path into the above kernel, restart our jupyter lab, and boom! We can now use our R env created in Rstudio in jupyter with no conda install quirks!