Behavior of RCTD multi-cores

[WuRAFY]

Firstly, thanks for the great tool. When I run RCTD with max_cores=8, I found that RCTD initiated 8 jobs and each job occupied multiple cores. Actually, RCTD occupied all cores available (50 in taotal) even I limit the max number cores. Can you help?

Much thanks, FY

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

other attached packages:
[1] sp_2.1-3           SeuratObject_4.1.0 Seurat_4.1.1       spacexr_2.2.1     

loaded via a namespace (and not attached):
  [1] nlme_3.1-153          spatstat.sparse_3.0-2 matrixStats_0.61.0   
  [4] RcppAnnoy_0.0.19      doParallel_1.0.16     RColorBrewer_1.1-3   
  [7] httr_1.4.7            sctransform_0.3.3     tools_4.1.0          
 [10] utf8_1.2.4            R6_2.5.1              irlba_2.3.5          
 [13] rpart_4.1-15          KernSmooth_2.23-20    uwot_0.1.11          
 [16] mgcv_1.8-38           rgeos_0.5-8           lazyeval_0.2.2       
 [19] colorspace_2.1-0      tidyselect_1.2.0      gridExtra_2.3        
 [22] compiler_4.1.0        progressr_0.9.0       cli_3.6.2            
 [25] plotly_4.10.0         scales_1.3.0          lmtest_0.9-39        
 [28] spatstat.data_3.0-1   ggridges_0.5.3        pbapply_1.5-0        
 [31] goftest_1.2-3         stringr_1.5.1         digest_0.6.34        
 [34] spatstat.utils_3.0-3  pkgconfig_2.0.3       htmltools_0.5.7      
 [37] parallelly_1.36.0     fastmap_1.1.1         htmlwidgets_1.5.4    
 [40] rlang_1.1.3           shiny_1.8.0           generics_0.1.2       
 [43] zoo_1.8-9             jsonlite_1.8.8        spatstat.random_3.1-5
 [46] ica_1.0-2             dplyr_1.1.4           magrittr_2.0.3       
 [49] patchwork_1.1.1       Matrix_1.3-4          Rcpp_1.0.8.3         
 [52] munsell_0.5.0         fansi_1.0.6           abind_1.4-5          
 [55] reticulate_1.24       lifecycle_1.0.4       stringi_1.8.3        
 [58] MASS_7.3-57           Rtsne_0.15            plyr_1.8.6           
 [61] grid_4.1.0            parallel_4.1.0        listenv_0.8.0        
 [64] promises_1.2.1        ggrepel_0.9.1         deldir_1.0-6         
 [67] miniUI_0.1.1.1        lattice_0.20-45       cowplot_1.1.1        
 [70] splines_4.1.0         tensor_1.5            pillar_1.9.0         
 [73] igraph_1.2.11         spatstat.geom_3.2-4   future.apply_1.8.1   
 [76] reshape2_1.4.4        codetools_0.2-18      leiden_0.3.9         
 [79] glue_1.7.0            data.table_1.14.2     png_0.1-8            
 [82] vctrs_0.6.5           httpuv_1.6.14         foreach_1.5.1        
 [85] polyclip_1.10-0       gtable_0.3.4          RANN_2.6.1           
 [88] purrr_1.0.2           spatstat.core_2.4-0   tidyr_1.3.1          
 [91] scattermore_0.8       future_1.33.1         ggplot2_3.4.4        
 [94] mime_0.12             xtable_1.8-4          later_1.3.2          
 [97] survival_3.2-13       viridisLite_0.4.2     tibble_3.2.1         
[100] iterators_1.0.13      cluster_2.1.2         globals_0.16.2       
[103] fitdistrplus_1.1-6    ellipsis_0.3.2        ROCR_1.0-11   

[ZhouJY-code]

Did you solve this problem? I’m facing the same issue.

[WuRAFY]

Did you solve this problem? I’m facing the same issue.

Sorry but nope, I got no idea why this would happen. But the output is good.

[WuRAFY]

Did you solve this problem? I’m facing the same issue.

The following code works for me, add it to head of your R script

library(RhpcBLASctl)
Sys.setenv(OMP_NUM_THREADS = "1", MKL_NUM_THREADS = "1",
           OPENBLAS_NUM_THREADS = "1", NUMEXPR_NUM_THREADS = "1", R_MAX_NUM_THREADS = "1")
blas_set_num_threads(1)
omp_set_num_threads(1)