Error in array(STATS, dims[perm]) : the length of 'dims' cannot be zero

[gfhtj]

Hi,
when I was trying to perform imputation with
imputed_data_bysaver <- saver(zerodata, ncores = 12, estimates.only = TRUE),
I got an error as below.
Error in array(STATS, dims[perm]) : the length of 'dims' cannot be zero
and there is a warning :
Warning messages:
1: In clean.data(x) : Make sure x is numeric.
2: In sweep(x[good.genes, ] + 1, 2, sf, "/") :
The range of STATS greater than'dim (x) [MARGIN] '

Can you please help me to fix this problem?
(when I perform this code on other dataset ,it works well,but fail in this one and encounter the error )

[mohuangx]

Can you run str on your zerodata and provide the output? You should make sure zerodata is a matrix with no NAs.

[gfhtj]

Yes, there are no NAs in my results, my zero is a dataframe ,

but it reported the aforementioned error. Additionally, when I tried to convert zerodata into a matrix, it reported another error:Error in if (!(grepl("matrix", class(x), ignore.case = TRUE))) { :
the condition has length > 1

[mohuangx]

Please try installing the Github version and try again after converting into a matrix.

Also, note that genes should be along the rows and cells should be along the columns.