cannot coerce type 'environment' to vector of type 'character'

[ynanli]

Hi,
Thanks for the nice package.

I am encounter issue when reading an AnnData object from h5ad file. Not sure what exactly the issue is.

> obj <- anndataR::read_h5ad(h5ad_path)
Error in as.vector(x, "character") : 
  cannot coerce type 'environment' to vector of type 'character'

I checked the h5ad file, and it seems to be fine.

> ad <- import("anndata", convert = FALSE)
> adata <- ad$read_h5ad(h5ad_path)
> adata
AnnData object with n_obs × n_vars = 403759 × 30647
    obs: 'sampleID', 'study', 'diagnosis', 'Disease_study', 'compartment', 'cell type', 'cell subtype', 'n_genes', 'log1p_n_genes_by_counts', 'total_counts', 'pct_counts_mitochondrial', 'Source'
    var: 'gene_ids', 'gene_name', 'mitochondrial'
    uns: 'cell subtype_colors', 'cell type_colors', 'compartment_colors', 'diagnosis_colors', 'neighbors', 'study_colors', 'umap'
    obsm: 'X_scVI', 'X_umap'
    layers: 'counts', 'log10_10k_normalized', 'log1p'
    obsp: 'connectivities', 'distances'

Any idea what might cause this issue?

> sessionInfo()
R version 4.4.2 (2024-10-31)
Platform: x86_64-pc-linux-gnu
Running under: Ubuntu 24.04.1 LTS

Matrix products: default
BLAS:   /usr/lib/x86_64-linux-gnu/blas/libblas.so.3.12.0 
LAPACK: /usr/lib/x86_64-linux-gnu/lapack/liblapack.so.3.12.0

locale:
 [1] LC_CTYPE=C.UTF-8       LC_NUMERIC=C           LC_TIME=C.UTF-8        LC_COLLATE=C.UTF-8    
 [5] LC_MONETARY=C.UTF-8    LC_MESSAGES=C.UTF-8    LC_PAPER=C.UTF-8       LC_NAME=C             
 [9] LC_ADDRESS=C           LC_TELEPHONE=C         LC_MEASUREMENT=C.UTF-8 LC_IDENTIFICATION=C   

time zone: Etc/UTC
tzcode source: system (glibc)

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

other attached packages:
[1] reticulate_1.42.0

loaded via a namespace (and not attached):
 [1] vctrs_0.6.5         cli_3.6.4           knitr_1.50          rlang_1.1.6         xfun_0.51          
 [6] purrr_1.0.4         png_0.1-8           jsonlite_1.9.1      bit_4.6.0           htmltools_0.5.8.1  
[11] rsconnect_1.3.4     rmarkdown_2.29      rappdirs_0.3.3      grid_4.4.2          evaluate_1.0.3     
[16] fastmap_1.2.0       yaml_2.3.10         lifecycle_1.0.4     compiler_4.4.2      Rcpp_1.0.14        
[21] rstudioapi_0.17.1   anndataR_0.1.0.9009 lattice_0.22-7      digest_0.6.37       R6_2.6.1           
[26] magrittr_2.0.3      Matrix_1.7-3        hdf5r_1.3.12        tools_4.4.2         withr_3.0.2        
[31] bit64_4.6.0-1    

> py_config()
python:         /mnt/data/home/yinanli/miniforge3/envs/popv/bin/python
libpython:      /mnt/data/home/yinanli/miniforge3/envs/popv/lib/libpython3.12.so
pythonhome:     /mnt/data/home/yinanli/miniforge3/envs/popv:/mnt/data/home/yinanli/miniforge3/envs/popv
version:        3.12.3 | packaged by conda-forge | (main, Apr 15 2024, 18:38:13) [GCC 12.3.0]
numpy:          /mnt/data/home/yinanli/miniforge3/envs/popv/lib/python3.12/site-packages/numpy
numpy_version:  2.1.3

NOTE: Python version was forced by use_python() function

[lazappi]

Hmmm...I'm not sure exactly. Could you please post the output from traceback() to help with working out where the error is coming from?

[ynanli]

@lazappi thanks for the hint. It seems like the h5ad was created using older anndata version.

> traceback()
36: as.character.default(x)
35: as.character(x)
34: grepl("\a|\033\\\\", txt)
33: make_link_file(txt)
32: make_link(x, type = "file")
31: transform(x)
30: inline_generic(app, val, style = style)
29: collapse_both_ends(x, sep, sep2, last, trunc, width, ellipsis)
28: ansi_collapse(x, sep = sep, sep2 = sep2, last = last, trunc = trunc, 
        style = col_style)
27: inline_collapse(inline_generic(app, val, style = style), style = style)
26: .transformer(expr, .envir)
25: .transformer(expr, .envir) %||% character()
24: (function (expr) 
    {
        eval_func <- as.character(.transformer(expr, .envir) %||% 
            character())
    })("v5")
23: glue(text, .envir = envir, .transformer = inline_transformer, 
        .open = paste0("<", envir$marker), .close = paste0(envir$marker, 
            ">"))
22: .transformer(expr, .envir)
21: .transformer(expr, .envir) %||% character()
20: (function (expr) 
    {
        eval_func <- as.character(.transformer(expr, .envir) %||% 
            character())
    })(".file <FCkNXbE-2v5FCkNXbE-2>")
19: glue(t$str, .envir = t$values, .transformer = inline_transformer, 
        .open = paste0("<", t$values$marker), .close = paste0(t$values$marker, 
            ">"), .trim = FALSE)
18: FUN(X[[i]], ...)
17: lapply(texts, function(t) {
        t$values$app <- app
        glue(t$str, .envir = t$values, .transformer = inline_transformer, 
            .open = paste0("<", t$values$marker), .close = paste0(t$values$marker, 
                ">"), .trim = FALSE)
    })
16: clii__inline(app, text, .list = .list)
15: app$inline(text, .list = .list)
14: clii__xtext(app, text, .list = .list, indent = indent, padding = padding, 
        wrap = wrap)
13: app$xtext(text, wrap = FALSE)
12: clii_inline_text(app, text)
11: (function (text) 
    clii_inline_text(app, text))(text = structure(list(str = "File <FCkNXbE-2.file <FCkNXbE-2v5FCkNXbE-2>FCkNXbE-2> is not a valid H5AD file.", 
        values = <environment>), class = "cli_glue_delay"))
10: do.call(app[[msg$type]], msg$args)
9: cli__fmt(rec, collapse, strip_newline)
8: cli_fmt(fun(..., .envir = .envir), collapse = collapse, strip_newline = TRUE)
7: FUN(X[[i]], ...)
6: vapply(X, FUN, FUN.VALUE = character(1), ..., USE.NAMES = USE.NAMES)
5: vcapply(message, format_inline, .envir = .envir)
4: cli_abort(c("File {.file {file}} is not a valid H5AD file.", 
       i = "Either the file is not an H5AD file or it was created with {.pkg anndata<0.8.0}."))
3: initialize(...)
2: HDF5AnnData$new(path, mode = mode)
1: anndataR::read_h5ad(h5ad_path)

The issue was ressoved by re-saving the h5ad using newer anndata version.

> ad <- import("anndata", convert = FALSE)
> adata <- ad$read_h5ad(h5ad_path)
> adata$write_h5ad(test_path)
> obj <- anndataR::read_h5ad(test_path, as = 'Seurat')
> obj
An object of class Seurat 
30647 features across 403759 samples within 1 assay 
Active assay: RNA (30647 features, 0 variable features)
 4 layers present: counts, log10_10k_normalized, log1p, X
 2 dimensional reductions calculated: X_scVI, X_umap

But I believe this is rather a tidious workaround?

[lazappi]

We currently only support H5AD files written with anndata>=0.8.0 which is more than 3 years old. This is because that was the first version which used a clearly defined on-disk format, prior to that the format was not standard enough to allow interoperability between languages (See #134).

You should have got a clearer error message though, so that's something we need to fix.