How to Melt R Data.Frame and Plot Group by Bar Plot

How to melt R data.frame and plot group by bar plot

Try:

mm <- melt(ddf, id='group')
ggplot(data = mm, aes(x = group, y = value, fill = variable)) + 
       geom_bar(stat = 'identity', position = 'dodge')

or

ggplot(data = mm, aes(x = group, y = value, fill = variable)) + 
       # `geom_col()` uses `stat_identity()`: it leaves the data as is.
       geom_col(position = 'dodge')

Frequency data frame for grouped bar plot

If you don't mind using data.table...

# generate your sample data
require(data.table)
dt <- data.table(structure(list(
Factor = c(0L, 1L, 0L, 1L, 0L, 1L, 1L, 0L, 1L, 0L, 0L, 1L, 0L, 1L, 0L, 
1L, 1L, 0L, 0L), VOR2Binned = c(3L, 3L, 3L, 3L, 2L, 2L, 3L, 2L, 3L, 2L, 
3L, 3L, 3L, 3L, 3L, 3L, 2L, 3L, 0L)), .Names = c("Factor","VOR2Binned"), 
row.names = c(NA, -19L), class = c("data.table", "data.frame")))

# count occurrences for each VOR2Binned
dt[, Frequency := .N, by=.(VOR2Binned, Factor)]

# let's make sure column Factor is really a factor
dt$Factor <- as.factor(dt$Factor)
# change name to Used and Available
levels(dt$Factor) <- c("Used", "Available")

# let's ggplot it!
ggplot(dt) + geom_col(aes(VOR2Binned, Frequency, fill=Factor),
                       position="dodge")

Reshape/melt a dataframe for scatter plot

So what did not work with melt? And what Problems did you have with geom_point()? Hard to say if this is what you want:

library( "reshape2" )
library( "ggplot2" )

df <- data.frame( x = rnorm(20), ya = rnorm(20), yb = rnorm(20), yc = rnorm(20) )
df <- melt(df, id.vars="x", variable.name="class", value.name="y")

ggplot( df, aes( x = x, y = y) ) +
  geom_point( aes(colour = class) )

ggplot( df, aes( x = x, y = y) ) +
  geom_point() +
  facet_wrap( "class" )

Plot fitted data from data frame as side-by-side barplot

Using base graphics:

# convert the one-row data frame to a two-row matrix
m <- matrix(unlist(df[2, ]), nrow = 2, byrow = TRUE)

# plot
barplot(m, beside = TRUE, col = c("blue", "red"), names.arg = seq_len(ncol(m)))

Possibly add a legend:

legend("topright", legend = c("UT", "TR"), fill = c("blue", "red"))

Showing bar plot side by side for two variables in dataframe

We may use base R barplot

barplot(t(`row.names<-`(as.matrix(df1[-1]), df1$Age_group)),
      legend = TRUE, beside = TRUE)

-output

---

Or if we wanted to use ggplot, reshape to 'long' format

library(ggplot2)
library(dplyr)
library(tidyr)
df1 %>%
    pivot_longer(cols = -Age_group) %>%
    ggplot(aes(x = Age_group, y = value, fill = name)) +
      geom_col(position = 'dodge') +
      theme_bw()

-output

How to Make a Grouped Barplot for a Factor with Many Levels

Assuming you want to plot means of Canopy_Index for each Under_Open, Topography cell, you can form means first:

df.means <- aggregate(Canopy_Index ~ Under_Open + Topography, df.melt, mean)

Then, plot df.means using the code from your question:

ggplot(df.means, aes(x=Topography, y=Canopy_Index, fill=Under_Open)) +
  geom_bar(stat="identity", position="dodge") +
  scale_fill_discrete(name="Canopy Type",
         labels=c("Under_tree"="Under Canopy", "Open_Canopy"="Open Canopy")) +
  xlab("Topographical Feature") + ylab("Canopy Index")

Result:

The reason why the bars are currently almost all of the same height is that you overlay multiple values per cell (as pointed out in the comments by Marijn Stevering), effectively plotting the max:

df.max <- aggregate(Canopy_Index ~ Under_Open + Topography, df.melt, max)
# Under_Open         Topography Canopy_Index
# 1   Under_tree Artificial_Surface           75
# 2  Open_Canopy Artificial_Surface           95
# 3   Under_tree          Bare_soil           95
# 4  Open_Canopy          Bare_soil           95
# 5   Under_tree              Grass           95
# 6  Open_Canopy              Grass           95
# 7   Under_tree             Litter           95
# 8  Open_Canopy             Litter           95
# 9   Under_tree        Undergrowth           95
# 10 Open_Canopy        Undergrowth           95

---

Related Topics