Misinformation when using anndataToGiotto with non default env_name

[wwang-chcn]

Describe the Error

The function anndataToGiotto, even though it runs successfully, throws an error message.
This will happen if I use an alternative env_name.

...

Error Message

# ERROR: The requested version of Python
# ('/Users/wwang/miniconda3/envs/giotto_env/bin/python')
# cannot be used, as another version of Python
# ('/Users/wwang/miniconda3/envs/giotto_ontrac_env/bin/python')
# has already been initialized. Please restart the R
# session if you need to attach reticulate to a different
# version of Python.

To Reproduce

Steps to reproduce the behavior, i.e.:

1. Install a custom conda env

installGiottoEnvironment(
  packages_to_install = c(
    "pandas==1.5.1",
    "networkx==2.8.8",
    "python-igraph==0.10.2",
    "leidenalg==0.9.0",
    "python-louvain==0.16",
    "python.app==1.4",
    "scikit-learn==1.1.3",
    "smfishhmrf",
    "scanpy",
    "session-info"
  ),
  pip_packages = c("python-louvain", "smfishhmrf", "session-info", "scanpy"),
  envname = "giotto_alt_env"
)

2. Try to read an .h5ad obj using anndataToGiotto

set_giotto_python_path("giotto_alt_env")
# a python path has been provided
# Using python path:
#  "/Users/wwang/miniconda3/envs/giotto_alt_env/bin/python"

giotto_all_slices_obj <- anndataToGiotto(file.path("data/03_session5/counts.h5ad"),
                                         env_name = "giotto_alt_env")
# python already initialized in this session
#  active environment : 'giotto_alt_env'
#  python version : 3.10
# checking default envname 'giotto_env'
# a system default python environment was found
# Using python path:
#  "/Users/wwang/miniconda3/envs/giotto_env/bin/python"
# ERROR: The requested version of Python
# ('/Users/wwang/miniconda3/envs/giotto_env/bin/python') cannot be used, as
# another version of Python
# ('/Users/wwang/miniconda3/envs/giotto_alt_env/bin/python') has already been
# initialized. Please restart the R session if you need to attach reticulate
# to a different version of Python.
# There are non numeric or integer columns for the spatial location
#  input at column position(s): 1
#  The first non-numeric column will be considered as a cell ID
#  to test for consistency with the expression matrix
#  Other non numeric columns will be removed
# 
# List item [1]:
#  spat_unit: cell
#  feat_type: rna
# > Cell metadata for spat_unit " cell " and feat_type " rna " already exists and
#  will be replaced with new metadata.
# Setting cell metadata [cell][rna]
# list depth of 2
# 
# List item [1]:
#  spat_unit: cell
#  feat_type: rna
# > Feat metadata for spat_unit " cell " and feat_type " rna " already exists and
#  will be replaced with new metadata.
# Setting feature metadata [cell][rna]
# 
# AnnData object successfully converted to Giotto.
# Warning message:
# Python is required for full Giotto functionality. 
# Spatial locations were not found. If spatial locations should have been found,
#         please modify the anndata object to include a keyword-value pair within the obsm slot,
#         in which the keyword contains the phrase "spatial" and the value corresponds to the spatial locations.
# 
#         In the Giotto Object resulting from this conversion, dummy locations will be used.

Expected behavior

This function ran successfully, no error message should thros.

Screenshots

System Information

R version 4.4.0 (2024-04-24)
Platform: aarch64-apple-darwin20
Running under: macOS Ventura 13.6.6

Matrix products: default
BLAS:   /System/Library/Frameworks/Accelerate.framework/Versions/A/Frameworks/vecLib.framework/Versions/A/libBLAS.dylib 
LAPACK: /Library/Frameworks/R.framework/Versions/4.4-arm64/Resources/lib/libRlapack.dylib;  LAPACK version 3.12.0

locale:
[1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8

time zone: America/New_York
tzcode source: internal

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base     

other attached packages:
[1] GiottoClass_0.3.5

loaded via a namespace (and not attached):
 [1] Matrix_1.7-0       jsonlite_1.8.8     compiler_4.4.0     gtools_3.9.5       Rcpp_1.0.13        magick_2.8.4       parallel_4.4.0     png_0.1-8         
 [9] yaml_2.3.8         fastmap_1.2.0      reticulate_1.38.0  lattice_0.22-6     deldir_2.0-4       dbscan_1.2-0       generics_0.1.3     igraph_2.0.3      
[17] knitr_1.47         backports_1.5.0    checkmate_2.3.2    R.utils_2.12.3     GiottoUtils_0.1.11 rlang_1.1.4        sp_2.1-4           terra_1.7-78      
[25] xfun_0.44          cli_3.6.3          withr_3.0.1        magrittr_2.0.3     digest_0.6.35      grid_4.4.0         rstudioapi_0.16.0  lifecycle_1.0.4   
[33] R.methodsS3_1.8.2  R.oo_1.26.0        evaluate_0.24.0    glue_1.7.0         data.table_1.15.4  codetools_0.2-20   rmarkdown_2.27     matrixStats_1.3.0 
[41] tools_4.4.0        pkgconfig_2.0.3    htmltools_0.5.8.1 

[wwang-chcn]

When I examined the code within the function anndataToGiotto, I found that the function didn't seem to be using the parameter env_name.

[jiajic]

It looks like the only thing that controls which python is used is the python_path param in there right now. But that one only passes the default python_path = NULL to createGiottoInstructions() in an attempt to initialize an environment following defaults.

set_giotto_python_path() has already done that for your specified environment.

I'm more surprised that the error from reticulate didn't stop execution.

[wwang-chcn]

We can do nothing if the error message yield from reticulate :(

[jiajic]

Well it's a good error to have and makes the user aware. We just need to remove the env_name param and use the python_path param since they're essentially handled the same way by Giotto, or something similar.

You can pass either an envname or a full path to a python binary to python_path param.

Looks like the reason it didn't stop execution is that somewhere in our implementation with createGiottoInstructions() we've used a try()

> createGiottoInstructions(python_path = "r-reticulate")
python already initialized in this session
 active environment : 'giotto_env'
 python version : 3.10
a python path has been provided
Using python path:
 "/projectnb/rd-spat/HOME/george/.conda/envs/r-reticulate/bin/python" Using
 python path:
 "/usr3/graduate/jiajic/.local/share/r-miniconda/envs/r-reticulate/bin/python"
$python_path
[1] "Error in required && !file_test(\"-f\", python) : \n  'length = 2' in coercion to 'logical(1)'\n"
attr(,"class")
[1] "try-error"
attr(,"condition")
<simpleError in required && !file_test("-f", python): 'length = 2' in coercion to 'logical(1)'>

[wwang-chcn]

Just test it. It won't throw any error message when I provide env name to python_path.