Problems with sum() function inside summarise using plyr and tidyverse!

Question

Hello, good afternoon!

I’m finding it difficult to use tidyverse, to perform a stacking function + summarise(soma=sum(value)) + pop = base%>%pivot_longer(cols=(),names_to(),values_to())%>%group_by()%>%summarise(soma=sum(value))%>%spread(variable,m). When using this code by modifying only the function within summarise(), of sum() for mean() or sd(), always worked, now to sum(), does not work, it does not sum, just repeats the value.

To check, I used reshape2::dcast(plyr::ddply(reshape2::melt())) that was the function I always had the habit of using, in addition to confronting the function sum() applied to the values in a subset() determined.

I would like to help you to check how to use by tidyverse, preferably without those exits from "summarise() ... (override with .groups argument)".

Below follows the base, and the codes:

  base<-data.frame(expand.grid(FAT1=1:4,FAT2=1:2,FAT3=1:2,AVA=1:6,REP=1:3),
VAR1=runif(3*96,0,1),
VAR2=runif(3*96,-1.5,0),
VAR3=runif(3*96,0,2.5))

require(reshape2)
require(plyr)
require(tidyverse)

mbase<-base %>% pivot_longer(cols=all_of(c("VAR1","VAR2","VAR3")),names_to = "variable",values_to = "value")




dcast(ddply(mbase, .(FAT1,FAT2,FAT3,AVA,variable), summarise, soma=sum(value)),FAT1*FAT2*FAT3*AVA~variable,fun.aggregate = sum,value.var ="soma" )
# (FAT1=1,FAT2=1,FAT3=1,AVA=1)$VAR1 = 1.704

sum(subset(base,FAT1==1 & FAT2==1 & FAT3==1 & AVA==1)[,"VAR1"])
# = 1.704



base %>% pivot_longer(cols=all_of(c("VAR1","VAR2","VAR3")),names_to = "variable",values_to = "value")%>%
  group_by(FAT1,FAT2,FAT3,AVA,variable) %>%  summarise(soma = sum(value)) %>% spread(variable,soma)

Answer 1

These two examples give different values of the sums in each row. The first example is the question, only with ungroup at the end in order to obtain TRUE when the results are compared with the second method, much simpler.

As for the second question, not to have the messages from summarise follows what is in the documentation, help('summarise'),

When .groups is not specified, you either get "drop_last" when all the Results are size 1, or "keep" if the size Varies. In addition, a message informs you of that Choice, unless the option "dplyr.summarise.inform" is set to FALSE.

op <- options(dplyr.summarise.inform = FALSE)

mbase <- base %>% 
  pivot_longer(
    cols = all_of(c("VAR1", "VAR2", "VAR3")),
    names_to = "variable",
    values_to = "value"
  ) %>% 
  group_by(FAT1, FAT2, FAT3, AVA, variable) %>% 
  summarise(soma = sum(value)) %>% 
  spread(variable, soma) %>%
  ungroup()


mbase2 <- base %>% 
  group_by(FAT1, FAT2, FAT3, AVA) %>% 
  summarise_at(vars(starts_with("VAR")), sum) %>%
  ungroup()


identical(mbase, mbase2)
#[1] TRUE