Dear @konsolerr,
Your tool is giving some good results but, I am getting an error when trying to calculate the significance of correlations using the lo$calculateSignificanceLevel(100) function.
My input data consisted of TPM values for 8000 genes and 1000 samples, to which I ran the following commands:
lo <- LinseedObject$new(a, topGenes = 5000); lo$calculatePairwiseLinearity(); lo$calculateSpearmanCorrelation(); lo$calculateSignificanceLevel(100);
However the last command raises the following error
Error in sparseMatrix(i = ij[, 1], j = ij[, 2], x = values, dims = c(genes, : length(x) must not exceed length(i)
I tried to do some debugging, and identified the problematic part of the code
ij <- which(upper.tri(self$pairwise) & self$pairwise > 0 & self$spearman > spearmanThreshold, arr.ind = T)
values <- self$pairwise[upper.tri(self$pairwise) & self$pairwise > 0 & self$spearman > spearmanThreshold] * self$spearman[upper.tri(self$pairwise) & self$pairwise > 0 & self$spearman > spearmanThreshold]
genes <- nrow(self$pairwise)
edges <- length(values)
So in this section of code, for my data, the variable ij is a matrix of indices with 163623 rows and 2 columns. However, the variable named as values contained 163630 values. Therefore, in next step, when trying to create a sparse matrix with the following code
sparsePP <- sparseMatrix(i = ij[, 1], j = ij[, 2], x = values, dims=c(genes, genes))
It raises an error saying
Error in sparseMatrix(i = ij[, 1], j = ij[, 2], x = values, dims = c(genes, : length(x) must not exceed length(i)
I believe this results from the facts that there is mismatch between the dimensions of ij and length of variable values.
Could you please take a look and help me trouble shoot it?
Dear @konsolerr,
Your tool is giving some good results but, I am getting an error when trying to calculate the significance of correlations using the lo$calculateSignificanceLevel(100) function.
My input data consisted of TPM values for 8000 genes and 1000 samples, to which I ran the following commands:
lo <- LinseedObject$new(a, topGenes = 5000); lo$calculatePairwiseLinearity(); lo$calculateSpearmanCorrelation(); lo$calculateSignificanceLevel(100);However the last command raises the following error
Error in sparseMatrix(i = ij[, 1], j = ij[, 2], x = values, dims = c(genes, : length(x) must not exceed length(i)I tried to do some debugging, and identified the problematic part of the code
ij <- which(upper.tri(self$pairwise) & self$pairwise > 0 & self$spearman > spearmanThreshold, arr.ind = T)values <- self$pairwise[upper.tri(self$pairwise) & self$pairwise > 0 & self$spearman > spearmanThreshold] * self$spearman[upper.tri(self$pairwise) & self$pairwise > 0 & self$spearman > spearmanThreshold]genes <- nrow(self$pairwise)edges <- length(values)So in this section of code, for my data, the variable ij is a matrix of indices with 163623 rows and 2 columns. However, the variable named as values contained 163630 values. Therefore, in next step, when trying to create a sparse matrix with the following code
sparsePP <- sparseMatrix(i = ij[, 1], j = ij[, 2], x = values, dims=c(genes, genes))It raises an error saying
Error in sparseMatrix(i = ij[, 1], j = ij[, 2], x = values, dims = c(genes, : length(x) must not exceed length(i)I believe this results from the facts that there is mismatch between the dimensions of ij and length of variable values.
Could you please take a look and help me trouble shoot it?