hostnamectl
# ...
# Operating System: CentOS Linux 7 (Core)
# CPE OS Name: cpe:/o:centos:centos:7
# Kernel: Linux 3.10.0-957.el7.x86_64
# Architecture: x86-64
# ...
conda --version
# conda 24.1.2
conda config --add channels bioconda
conda config --add channels conda-forge
conda config --show channels
# channels:
# - conda-forge
# - bioconda
# - defaults
#
# create the environment
conda activate base
conda create --name publish --yes
conda activate publish
conda install --name publish --file require_publish --yes
pip install torch==2.0.1 torchvision==0.15.2 torchaudio==2.0.2 --index-url https://download.pytorch.org/whl/cpu
pip install samap
pip install came
conda list --name publish | grep -E "^(python )|(scanpy)|(torch)|(r-base )|(r-seurat)|(r-seuratobject)|(came)|(^samap)"
# python 3.9.18 h0755675_1_cpython conda-forge
# pytorch 2.1.2 cpu_generic_py39h163b580_0 conda-forge
# scanpy 1.9.3 pyhd8ed1ab_0 conda-forge
# samap 1.0.15 pypi_0 pypi
# came 0.1.13 pypi_0 pypi
# r-base 4.3.3 hb8ee39d_0 conda-forge
# r-seurat 5.0.0 r43ha503ecb_0 conda-forge
# r-seuratobject 5.0.1 r43ha503ecb_0 conda-forge
# create jupyter kernel
conda activate publish
python -m ipykernel install --user --name 'publish' --display-name 'publish'
## [r-kernel](https://irkernel.github.io/installation/#linux-panel)
echo 'IRkernel::installspec(user = TRUE,name="R_publish",display="R_publish")' > temp_r_script.r
Rscript temp_r_script.r
rm temp_r_script.r
## check jupyter kernel
conda activate
jupyter kernelspec list
# Available kernels:
# ...
# publish ....local/share/jupyter/kernels/publish
# r_publish ....local/share/jupyter/kernels/r_publish
# ...
We provide a small dataset and tow scripts, demo.ipynb and demo_Seurat.ipynb to test if the environment is valid.