Issue with replacement length = 0?

[linnykos]

Hi Mengjie,

I have been using VIPER for a few days, and it's been great. However, I get this strange error when I try running it on data I downsample myself, and I was wondering if you had thoughts about what might be the issue or how to resolve it.

The following code works fine (1014 cells, 542 genes)

df <- read.csv("https://raw.githubusercontent.com/linnylin92/VIPER/master/data/singlecell.csv", row.names = 1)
res <- VIPER::VIPER(df, num = 5000, percentage.cutoff = 0.1, minbool = FALSE, alpha = 1,
                    report = FALSE, outdir = NULL, prefix = NULL)

However, the following code gives me an error (where df2 is now my downsampled variant of df)

df2 <- read.csv("https://raw.githubusercontent.com/linnylin92/VIPER/master/data/singlecell_downsample.csv", row.names = 1)
res <- VIPER::VIPER(df2, num = 5000, percentage.cutoff = 0.1, minbool = FALSE, alpha = 1,
                    report = FALSE, outdir = NULL, prefix = NULL)

The error I receive is Error in imputation_by_samples_posterior_expectation(data, selected_logxx, : Evaluation error: replacement has length zero. Is there an implicit requirement (or a documentation thing I missed) stating something such as the minimum number of zeros each gene/cell must have?

Thanks for your help!

[ChenMengjie]

The percentage of zero cutoff is specified using “percentage.cutoff = 0.1”. I think this error means you don’t have genes pass the cutoff. Maybe try sth like 0.5? -Mengjie On Mar 14, 2019, at 11:57 AM, linnylin92 <notifications@github.com<mailto:notifications@github.com>> wrote: Hi Mengjie, I have been using VIPER for a few days, and it's been great. However, I get this strange error when I try running it on data I downsample myself, and I was wondering if you had thoughts about what might be the issue or how to resolve it. The following code works fine (1014 cells, 542 genes) df <- read.csv("https://raw.githubusercontent.com/linnylin92/VIPER/master/data/singlecell.csv", row.names = 1) res <- VIPER::VIPER(df, num = 5000, percentage.cutoff = 0.1, minbool = FALSE, alpha = 1, report = FALSE, outdir = NULL, prefix = NULL) However, the following code gives me an error (where df2 is now my downsampled variant of df) df2 <- read.csv("https://raw.githubusercontent.com/linnylin92/VIPER/master/data/singlecell_downsample.csv", row.names = 1) res <- VIPER::VIPER(df2, num = 5000, percentage.cutoff = 0.1, minbool = FALSE, alpha = 1, report = FALSE, outdir = NULL, prefix = NULL) The error I receive is Error in imputation_by_samples_posterior_expectation(data, selected_logxx, : Evaluation error: replacement has length zero. Is there an implicit requirement (or a documentation thing I missed) stating something such as the minimum number of zeros each gene/cell must have? Thanks for your help! — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub<#2>, or mute the thread<https://github.com/notifications/unsubscribe-auth/AP417bOMKHexlf9opD1rbWzhoHq86HEHks5vWn9rgaJpZM4b0vtZ>.