How to Melt Data Frame

Question

I have a data frame, dratiodf:

   Date       D10       D21       D63      D126      D252
2019-09-23 0.1557585 0.3977892 0.8583822 0.7153107 0.7517688
2019-09-24 0.1513844 0.2142586 0.7501128 0.6736790 0.7275896
2019-09-25 0.5314460 0.4254800 0.8604258 0.6612713 0.7469207
2019-09-26 0.5135381 0.4250006 0.9424716 0.7008503 0.7661933
2019-09-27 0.4816461 0.2371428 0.7969672 0.6351504 0.7307106
2019-09-30 0.6414031 0.3407633 0.8207621 0.6854996 0.7346074

What I would like to do is melt the columns together to have a data frame that looks like this:

Date:        Type:  Value:
2019-09-23   D10    0.1557585
2019-09-23   D21    0.3977892
2019-09-23   D63    0.8583822
2019-09-23   D126   0.7153107
2019-09-23   D252   0.7517688
2019-09-34   D10    0.1513844
2019-09-34   D21    0.2142586

I want this so that I can facet the eventual plot by Type, like this:

ggplot(dratiodf, aes(x=Date, y=Value))+
 geom_line()+
 facet_wrap(.~type)+
 theme_wsj()

I've tried using the melt function, but I just can't wrap my head around how to use it.

Also, can you spot any problem with my graph code, something that won't work?

https://stackoverflow.com/questions/58193116/using-r-how-to-plot-and-color-with-multiple-facet — M--, Oct 01 '19 at 22:52

akrun · Answer 1 · 2019-10-01T23:01:32.110

We can use gather

library(tidyr) 
library(dplyr)
df1 %>%
  gather(Type, Value, -Date)

Or using pivot_longer (from the tidyr_1.0.0)

df1 %>%
  pivot_longer(cols = - Date, names_to =  "Type", values_to = "Value")
# A tibble: 30 x 3
#   Date       Type  Value
#   <chr>      <chr> <dbl>
# 1 2019-09-23 D10   0.156
# 2 2019-09-23 D21   0.398
# 3 2019-09-23 D63   0.858
# 4 2019-09-23 D126  0.715
# 5 2019-09-23 D252  0.752
# 6 2019-09-24 D10   0.151
# 7 2019-09-24 D21   0.214
# 8 2019-09-24 D63   0.750
# 9 2019-09-24 D126  0.674
#10 2019-09-24 D252  0.728
# … with 20 more rows

data

df1 <- structure(list(Date = c("2019-09-23", "2019-09-24", "2019-09-25", 
"2019-09-26", "2019-09-27", "2019-09-30"), D10 = c(0.1557585, 
0.1513844, 0.531446, 0.5135381, 0.4816461, 0.6414031), D21 = c(0.3977892, 
0.2142586, 0.42548, 0.4250006, 0.2371428, 0.3407633), D63 = c(0.8583822, 
0.7501128, 0.8604258, 0.9424716, 0.7969672, 0.8207621), D126 = c(0.7153107, 
0.673679, 0.6612713, 0.7008503, 0.6351504, 0.6854996), D252 = c(0.7517688, 
0.7275896, 0.7469207, 0.7661933, 0.7307106, 0.7346074)), 
 class = "data.frame", row.names = c(NA, 
-6L))

What does the '-' do, in '-Date' ran it without it and got a completely different answer — L.Zingg, Oct 01 '19 at 22:52

score 2 · Answer 2 · answered Oct 01 '19 at 23:04

Next time, provide a reproducible data frame like this.

df <- data.frame(Date = c("2019-09-23", "2019-09-24", "2019-09-25",
                          "2019-09-26", "2019-09-27", "2019-09-30"),
                 D10 = c(0.1557585, 0.1513844, 0.5314460,
                         0.5135381, 0.4816461, 0.6414031),
                 D21 = c(0.3977892, 0.2142586, 0.4254800, 
                         0.4250006, 0.2371428, 0.3407633),
                 D63 = c(0.8583822, 0.7501128, 0.8604258,
                         0.9424716, 0.7969672, 0.8207621),
                 D126 = c(0.7153107, 0.6736790, 0.6612713,
                          0.7008503, 0.6351504, 0.6854996),
                 D252 = c(0.7517688, 0.727589, 0.7469207,      
                          0.7661933, 0.7307106, 0.7346074))

Here's how you want to use the melt function.

library(reshape2)
dratiodf <- melt(df,
                 id.vars = "Date",
                 variable.name = "Type",
                 value.name = "Value")

Then your plot should work with the following.

ggplot(dratiodf, aes(x = Date, y = Value, group = Type)) +
    geom_line() +
    facet_wrap(~ Type) +
    theme_wsj() +
    theme(axis.text.x = element_text(angle = 45, hjust = 1))

How to Melt Data Frame

2 Answers2

data