Error in dtype(V) : could not find function "dtype"

[jluft-uoe]

I cannot get the core 'snmf' function to run in my environment.

Is there a dependency I'm missing?

My code:

data <- load_data(tnbc)
counts <- data$counts
S <- data$S
results <- snmf(gpu.matrix(counts), S, 5, niter=2000, tol=1e-4, num_initializations=10, probs=0.75, seed=42)

The error:

Error in dtype(V) : could not find function "dtype"

My session:

R version 4.6.1 (2026-06-24 ucrt)
Platform: x86_64-w64-mingw32/x64
Running under: Windows 11 x64 (build 26200)

Matrix products: default
  LAPACK version 3.12.1

locale:
[1] LC_COLLATE=English_United Kingdom.utf8  LC_CTYPE=English_United Kingdom.utf8    LC_MONETARY=English_United Kingdom.utf8
[4] LC_NUMERIC=C                            LC_TIME=English_United Kingdom.utf8    

time zone: Europe/London
tzcode source: internal

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

other attached packages:
[1] here_1.0.2         GPUmatrix_1.0.3    torch_0.17.0       Seurat_5.5.1       SeuratObject_5.4.0 sp_2.2-3           SNMF_0.0.1.1      

loaded via a namespace (and not attached):
  [1] deldir_2.0-4           pbapply_1.7-4          gridExtra_2.3.1        rlang_1.3.0            magrittr_2.0.5         RcppAnnoy_0.0.23      
  [7] otel_0.2.0             spatstat.geom_3.8-2    matrixStats_1.5.0      ggridges_0.5.7         compiler_4.6.1         callr_3.8.0           
 [13] png_0.1-9              vctrs_0.7.3            reshape2_1.4.5         stringr_1.6.0          pkgconfig_2.0.3        fastmap_1.2.0         
 [19] promises_1.5.0         bit_4.6.0              purrr_1.2.2            xfun_0.60              jsonlite_2.0.0         goftest_1.2-3         
 [25] later_1.4.8            spatstat.utils_3.2-4   irlba_2.3.7            parallel_4.6.1         cluster_2.1.8.2        R6_2.6.1              
 [31] ica_1.0-3              stringi_1.8.9          RColorBrewer_1.1-3     spatstat.data_3.1-9    reticulate_1.46.0      parallelly_1.48.0     
 [37] spatstat.univar_3.2-0  lmtest_0.9-40          scattermore_1.2        Rcpp_1.1.2             knitr_1.51             tensor_1.5.1          
 [43] future.apply_1.20.2    zoo_1.9-0              sctransform_0.4.3      httpuv_1.6.17          Matrix_1.7-5           splines_4.6.1         
 [49] igraph_2.3.3           tidyselect_1.2.1       rstudioapi_0.19.0      abind_1.4-8            yaml_2.3.12            spatstat.random_3.5-1 
 [55] codetools_0.2-20       miniUI_0.1.2           spatstat.explore_3.8-2 processx_3.9.0         listenv_1.0.0          lattice_0.22-9        
 [61] tibble_3.3.1           plyr_1.8.9             shiny_1.14.0           S7_0.2.2               ROCR_1.0-12            evaluate_1.0.5        
 [67] Rtsne_0.17             future_1.75.0          fastDummies_1.7.6      survival_3.8-6         polyclip_1.10-7        fitdistrplus_1.2-6    
 [73] pillar_1.11.1          BiocManager_1.30.27    KernSmooth_2.23-26     renv_1.2.2             plotly_4.12.1          generics_0.1.4        
 [79] rprojroot_2.1.1        RcppHNSW_0.7.0         ggplot2_4.0.3          scales_1.4.0           coro_1.1.0             globals_0.19.1        
 [85] xtable_1.8-8           glue_1.8.1             tools_4.6.1            data.table_1.18.4      RSpectra_0.16-2        RANN_2.6.2            
 [91] dotCall64_1.2          cowplot_1.2.0          grid_4.6.1             tidyr_1.3.2            nlme_3.1-169           patchwork_1.3.2       
 [97] cli_3.6.6              spatstat.sparse_3.2-0  spam_2.11-4            viridisLite_0.4.3      dplyr_1.2.1            uwot_0.2.4            
[103] gtable_0.3.6           digest_0.6.39          progressr_1.0.0        ggrepel_0.9.8          htmlwidgets_1.6.4      farver_2.1.2          
[109] htmltools_0.5.9        lifecycle_1.0.5        httr_1.4.8             mime_0.13              bit64_4.8.2            MASS_7.3-65    

[LuisAlonsoEsteban]

Thanks for reporting this.

Your session already includes GPUmatrix and torch. This error points to a namespace lookup bug in SNMF: it called dtype() without explicitly resolving it from GPUmatrix.

I’ve corrected this lookup and related namespace issues. Please restart R and reinstall the updated package:

remotes::install_github("LuisAlonsoEsteban/SNMF", force = TRUE)

Then pass the count matrix directly to snmf(); it handles the tensor conversion internally:

library(SNMF)

data(tnbc)
input <- load_data(tnbc)

results <- snmf(
    input$counts, input$S, k = 5,
    niter = 2000, tol = 1e-4,
    num_initializations = 10,
    probs = 0.75, seed = 42
)

Thank you for your interest in using SNMF.