search

whitlock / OutFLANK

A procedure to find Fst outliers based on an inferred distribution of neutral Fst
18 stars 9 forks source link

MakeDiploidFSTMat Error X 2 #7

Open dani-davenport opened 7 years ago

dani-davenport commented 7 years ago

Dear Micheal and Katie

Hoping you could take a look at this issue below and help me out. Got it working, then R crashed, now even after reinstalling the packages and updating still cant get it to work.

FSTmat<- MakeDiploidFSTMat(SNPmat=SNPmat, locusNames=locusNames, popNames=as.character(pops))

Error in tapply(snpDataTemp, list(popNameTemp, snpDataTemp), length) : arguments must have same length

So we try to transpose matrix... because of conflicting data input information between (a) OUTFLANK readme pdf and ?MakeDiploidFSTMat

SNPmat<- t(SNPmat)

FSTmat<- MakeDiploidFSTMat(SNPmat=SNPmat, locusNames=locusNames, popNames=as.character(pops)) [1] "Error: your population names do not match your SNP matrix" Error in MakeDiploidFSTMat(SNPmat = SNPmat, locusNames = locusNames, popNames = as.character(pops)) : no loop for break/next, jumping to top level

Please find attached my genotype data file and locusNames and pops OutFLANK_data.txt locusNames_OF.txt popNames_OF.txt

DrK-Lo commented 7 years ago

Hello,

OutFLANK uses a 9 for missing data, as described in the pdf. When I fixed your data, it worked just fine. That is because NAs are removed by the algorithm, resulting in a different length between your vectors.

library(OutFLANK) o <- read.table("OUTFLANK_data.txt") dim(o) p <- read.csv("popNames_OF.txt") head(p) l <- read.csv("locusNames_OF.txt") head(l)

Use 9 for missing data

sum(is.na(o)) # should not contain NAs o2 <- o o2[is.na(o2)] <- 9 sum(is.na(o2)) dim(o2) levels(factor(as.numeric(unlist(o2))))

out <- MakeDiploidFSTMat(SNPmat=o2, locusNames=l$x, popNames=p$x)

out <- MakeDiploidFSTMat(SNPmat=o, locusNames=l$x, popNames=p$x)

replicate error when use NA for missing data

On Aug 3, 2017, at 1:29 AM, tunaloveheart notifications@github.com wrote:

Dear Micheal and Katie

Hoping you could take a look at this issue below and help me out. Got it working, then R crashed, now even after reinstalling the packages and updating still cant get it to work.

FSTmat<- MakeDiploidFSTMat(SNPmat=SNPmat, locusNames=locusNames, popNames=as.character(pops))

Error in tapply(snpDataTemp, list(popNameTemp, snpDataTemp), length) : arguments must have same length

So we try to transpose matrix... because of conflicting data input information between (a) OUTFLANK readme pdf and ?MakeDiploidFSTMat

SNPmat<- t(SNPmat)

FSTmat<- MakeDiploidFSTMat(SNPmat=SNPmat, locusNames=locusNames, popNames=as.character(pops)) [1] "Error: your population names do not match your SNP matrix" Error in MakeDiploidFSTMat(SNPmat = SNPmat, locusNames = locusNames, popNames = as.character(pops)) : no loop for break/next, jumping to top level

Please find attached my genotype data file and locusNames and pops OutFLANK_data.txt https://github.com/whitlock/OutFLANK/files/1196071/OutFLANK_data.txt locusNames_OF.txt https://github.com/whitlock/OutFLANK/files/1196073/locusNames_OF.txt popNames_OF.txt https://github.com/whitlock/OutFLANK/files/1196074/popNames_OF.txt — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/whitlock/OutFLANK/issues/7, or mute the thread https://github.com/notifications/unsubscribe-auth/AGjUbYo7aJls_0dAODZBJhyLCDqagoQJks5sUVq6gaJpZM4Or-QX.

dani-davenport commented 7 years ago

Hey Katie

Thanks so much I knew it would be something stupid. When R crashed I mustn't have saved the right data frame which worked the first time because I change the NA to 9. Thanks again and sorry for my stupid question!!

Dani

Sent from my iPhone

On 3 Aug 2017, at 7:35 pm, Katie Lotterhos notifications@github.com wrote:

Hello,

OutFLANK uses a 9 for missing data, as described in the pdf. When I fixed your data, it worked just fine. That is because NAs are removed by the algorithm, resulting in a different length between your vectors.

library(OutFLANK) o <- read.table("OUTFLANK_data.txt") dim(o) p <- read.csv("popNames_OF.txt") head(p) l <- read.csv("locusNames_OF.txt") head(l)

Use 9 for missing data

sum(is.na(o)) # should not contain NAs o2 <- o o2[is.na(o2)] <- 9 sum(is.na(o2)) dim(o2) levels(factor(as.numeric(unlist(o2))))

out <- MakeDiploidFSTMat(SNPmat=o2, locusNames=l$x, popNames=p$x)

out <- MakeDiploidFSTMat(SNPmat=o, locusNames=l$x, popNames=p$x)

replicate error when use NA for missing data

On Aug 3, 2017, at 1:29 AM, tunaloveheart notifications@github.com wrote:

Dear Micheal and Katie

Hoping you could take a look at this issue below and help me out. Got it working, then R crashed, now even after reinstalling the packages and updating still cant get it to work.

FSTmat<- MakeDiploidFSTMat(SNPmat=SNPmat, locusNames=locusNames, popNames=as.character(pops))

Error in tapply(snpDataTemp, list(popNameTemp, snpDataTemp), length) : arguments must have same length

So we try to transpose matrix... because of conflicting data input information between (a) OUTFLANK readme pdf and ?MakeDiploidFSTMat

SNPmat<- t(SNPmat)

FSTmat<- MakeDiploidFSTMat(SNPmat=SNPmat, locusNames=locusNames, popNames=as.character(pops)) [1] "Error: your population names do not match your SNP matrix" Error in MakeDiploidFSTMat(SNPmat = SNPmat, locusNames = locusNames, popNames = as.character(pops)) : no loop for break/next, jumping to top level

Please find attached my genotype data file and locusNames and pops OutFLANK_data.txt https://github.com/whitlock/OutFLANK/files/1196071/OutFLANK_data.txt locusNames_OF.txt https://github.com/whitlock/OutFLANK/files/1196073/locusNames_OF.txt popNames_OF.txt https://github.com/whitlock/OutFLANK/files/1196074/popNames_OF.txt — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/whitlock/OutFLANK/issues/7, or mute the thread https://github.com/notifications/unsubscribe-auth/AGjUbYo7aJls_0dAODZBJhyLCDqagoQJks5sUVq6gaJpZM4Or-QX.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub, or mute the thread.

saarman commented 6 years ago

Thanks I had the same problem. using plink --recodeA gives a genotype matrix with NA, thus this problem. Could you place this conversion of formats [is.na(genotype)] <- 9 in the vignette so that people without a bunch of experience in R can fix their input files to match this non-standard format (i.e. 9 instead of NA)?