Geek Answers Handbook

R melt error (change) - number of rows

Look at this test data set with five observations and two factors.

test respid s2q1 s2q2 1 32 9 5 - V. satisfied 2 35 10 - Definitely 5 - V. satisfied 3 148 10 - Definitely 5 - V. satisfied 4 371 3 2 5 416 10 - Definitely 5 - V. satisfied

When I run the following melt command, I get an error message that does not make sense to me. Especially because I used the same commands before successfully executing.

 require(reshape2) qhist <- melt(test, id="respid") Error in data.frame(ids, variable, value, stringsAsFactors = FALSE) : arguments imply differing number of rows: 5, 10

Data has been imported using

 spss.get("data.sav", lowernames=T, use.value.labels=T,max.value.labels=13) dput(test) structure(list(respid = structure(c(32L, 35L, 148L, 371L, 416L), label = structure("Serial Number", .Names = "respid"), class = "labelled"), s2q1 = structure(c(10L, 11L, 11L, 4L, 11L), .Label = c("0 - Definitely not", "1", "2", "3", "4", "5", "6", "7", "8", "9", "10 - Definitely", "Don't know"), class = c("labelled", "factor"), label = structure("S2Q1 xxx?", .Names = "s2q1")), s2q2 = structure(c(5L, 5L, 5L, 2L, 5L), .Label = c("1 - Not at all satisfied", "2", "3", "4", "5 - V. satisfied", "Don't know"), class = c("labelled", "factor"), label = structure("S2Q2 xxx?", .Names = "s2q2"))), .Names = c("respid", "s2q1", "s2q2"), row.names = c(NA, 5L), class = "data.frame")
+4
source share
1 answer

This is caused by the respid structure, i.e. because its class is "tagged", but it works:

 library(reshape) reshape::melt(test, id="respid") respid variable value 1 32 s2q1 9 2 35 s2q1 10 - Definitely 3 148 s2q1 10 - Definitely 4 371 s2q1 3 5 416 s2q1 10 - Definitely 6 32 s2q2 5 - V. satisfied 7 35 s2q2 5 - V. satisfied 8 148 s2q2 5 - V. satisfied 9 371 s2q2 2 10 416 s2q2 5 - V. satisfied

or if you do not use reshape , then

 class(test$respid) <- "factor" reshape2::melt(test, id="respid")

works too.

+5
source

More articles:


All Articles