A list, a matrix and impute.knn for making data for SAMR
1
0
Entering edit mode
John Herbert ▴ 90
@john-herbert-3373
Last seen 7.1 years ago
Dear Bioconductors, Please can you help me with small annoying problem. I want to impute a matrix that has 4 columns of data. Using the impute.knn package returns me a list not a matrix. Using as.matrix to return the data returned from impute.knn does not give me the same matrix structure as I began with? It must be a simple answer? My code: library(impute); imputed_data <- impute.knn(normalised_data, k=5); > dim(normalised_data) [1] 43376 4 > dim(imputed_data) NULL > class(normalised_data) [1] "matrix" > class(imputed_data) [1] "list" Why does impute.knn not return the data as you would sensibly think it would, ie in the same format as your microarray data? Please tell me if I am missing something or the solution? Thanks a lot for any help. Kind regards, John. [[alternative HTML version deleted]]
Microarray impute Microarray impute • 808 views
0
Entering edit mode
@elin-axelsson-2940
Last seen 7.1 years ago
Hi John, if you look in ?impute.knn under the "value" section you will see that function returns 3 things; the data matrix, the rng.seed and the rng.state. It's returned as list, as you noticed, if you want the data matrix you simply write imputed_data$data. Hope this helps, Elin John Herbert wrote: > Dear Bioconductors, > > Please can you help me with small annoying problem. > > > > I want to impute a matrix that has 4 columns of data. Using the > impute.knn package returns me a list not a matrix. > > > > Using as.matrix to return the data returned from impute.knn does not > give me the same matrix structure as I began with? > > It must be a simple answer? > > > > My code: > > > > library(impute); > > imputed_data <- impute.knn(normalised_data, k=5); > > > > >> dim(normalised_data) >> > > [1] 43376 4 > > >> dim(imputed_data) >> > > NULL > > >> class(normalised_data) >> > > [1] "matrix" > > >> class(imputed_data) >> > > [1] "list" > > > > Why does impute.knn not return the data as you would sensibly think it > would, ie in the same format as your microarray data? > > Please tell me if I am missing something or the solution? > > > > Thanks a lot for any help. > > > > Kind regards, > > > > John. > > > [[alternative HTML version deleted]] > > _______________________________________________ > Bioconductor mailing list > Bioconductor at stat.math.ethz.ch > https://stat.ethz.ch/mailman/listinfo/bioconductor > Search the archives: http://news.gmane.org/gmane.science.biology.informatics.conductor > ADD COMMENT 0 Entering edit mode Hello Elin. Thanks very much for solving my problem, it helps a lot. I knew it would be simple, just like me :-). Actually, that is very useful advice as I was wondering if it told you somewhere what it returned. I should read more thoroughly before posting, sorry. Kind regards, John. -----Original Message----- From: Elin Axelsson [mailto:elin@ebi.ac.uk] Sent: 11 August 2009 13:06 To: John Herbert Cc: bioconductor at stat.math.ethz.ch Subject: Re: [BioC] A list, a matrix and impute.knn for making data for SAMR Hi John, if you look in ?impute.knn under the "value" section you will see that function returns 3 things; the data matrix, the rng.seed and the rng.state. It's returned as list, as you noticed, if you want the data matrix you simply write imputed_data$data. Hope this helps, Elin John Herbert wrote: > Dear Bioconductors, > > Please can you help me with small annoying problem. > > > > I want to impute a matrix that has 4 columns of data. Using the > impute.knn package returns me a list not a matrix. > > > > Using as.matrix to return the data returned from impute.knn does not > give me the same matrix structure as I began with? > > It must be a simple answer? > > > > My code: > > > > library(impute); > > imputed_data <- impute.knn(normalised_data, k=5); > > > > >> dim(normalised_data) >> > > [1] 43376 4 > > >> dim(imputed_data) >> > > NULL > > >> class(normalised_data) >> > > [1] "matrix" > > >> class(imputed_data) >> > > [1] "list" > > > > Why does impute.knn not return the data as you would sensibly think it > would, ie in the same format as your microarray data? > > Please tell me if I am missing something or the solution? > > > > Thanks a lot for any help. > > > > Kind regards, > > > > John. > > > [[alternative HTML version deleted]] > > _______________________________________________ > Bioconductor mailing list > Bioconductor at stat.math.ethz.ch > https://stat.ethz.ch/mailman/listinfo/bioconductor > Search the archives: http://news.gmane.org/gmane.science.biology.informatics.conductor >