Error in Beta[n, ] : subscript out of bounds

[houruijiejerry]

I encountered an error when running "obj <- Constr_net(obj)":
Constructing network for T cell
| | 0 % ~calculating Error in Beta[n, ] : subscript out of bounds

[Lee0498]

Hi, thanks for reporting the issue. To help us better understand and address the problem, we need more details.

Could you please provide the following:
1. The full code or a minimal reproducible example that triggers the error (including how you created the object and ran the workflow).
2. The dimensions of the data (dim(data)), and any relevant data processing steps you applied before the error occurred.
3. If possible, please include a small sample of your data.

Additionally, could you try running the workflow with parallelization disabled by setting n.cores = 1? I will check it later.

[houruijiejerry]

Thank you for your reply!
The problem has been solved by setting n.cores = 1.

However, I encountered another error when running the following code:

obj <- CreateScRank(input = scRNA,
species = 'human',
cell_type = 'detail_celltype',
target = 'FGFR1')
obj <- Constr_net(
obj,
select_ratio = 0.5,
n_selection = 10,
cut_ratio = 0.95,
keep_ratio = 0.25,
min_cells = 25,
n.core = 1,
n.core_cp = 1,
use_py = F,
env = "base"
)

Error in seq_len(n) : argument must be coercible to non-negative integer
Calls: rank_celltype -> lapply -> FUN -> sapply -> lapply
In addition: Warning messages:
1: In mclapply(argsList, FUN, mc.preschedule = preschedule, mc.set.seed = set.seed, :
scheduled cores 1, 2, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 16, 17, 18, 19, 20, 21, 27, 28, 31 did not deliver results, all values of the jobs will be affected
2: In seq_len(n) : first element used of 'length.out' argument

scRNA is a Seurat Object:

scRNA
An object of class Seurat
31884 features across 30171 samples within 1 assay
Active assay: RNA (31884 features, 2000 variable features)
3 layers present: counts, data, scale.data
3 dimensional reductions calculated: pca, harmony, umap

[Lee0498]

Through the information you provided, it seems that running in single-core mode have resolved the issue in the Constr_net() step. However, the new error appears to have occurred during the rank_celltype() step. Could you please try running the function 'rank_celltype()' again with single-core mode by setting n.core = 1?

obj <- rank_celltype(obj, n.core = 1)

This will help us determine whether the problem is related to parallel computation or to the data itself.

[houruijiejerry]

My apologies if my previous description wasn't clear. The new error occurred when running 'Constr_net' step.

[Lee0498]

Thanks for clarifying!

Since the error message includes

Error in seq_len(n) : argument must be coercible to non-negative integer
Calls: rank_celltype -> lapply -> FUN -> sapply -> lapply

it indicates that the rank_celltype() function was indeed called during the process. This makes it a bit tricky to pinpoint the exact cause without reproducing the issue locally.

To investigate further, could you please share a small subset of your data that can reproduce the problem?
You can send it via email (chengyuli@zju.edu.cn) if that’s more convenient.