CODEWITHSUNDEEP
rggplot2
KongBorup
KongBorup

Reputation: 13

Creating a bar plot with a single vector with ggplot

Creating a barplot from a single vector/column in a .csv file is no problem in vanilla R:

msgs <- read.csv(file = "C:/my/csv/file.csv", header = TRUE, sep = ",")

barplot(table(msgs$Author))

This code produces following barplot:

enter image description here

Though, I am having a hard time replicating this with the ggplot2 package. To do it with a qplot(), this code works:

qplot(as.factor(msgs$Author), geom="bar")

But I need it as a ggplot(), which I can't seem to make work. I have tried:

ggplot() + 
aes(table(msgs$Author)) + 
geom_bar(stat = "identity", width = .75, ill = "tomato3")

(which didn't work at all) and

ggplot(msgs, aes(x = Author, y = Timestamp)) +
geom_bar(stat = "identity", width = .75, fill = "tomato3")

The above snippet does show correct result with the bars

enter image description here

but the y-axis is all wrong. I don't want Timestamp to be the y-axis, but rather the amount of times every Author shows up in the vector/column.

I can't really find a way of providing ggplot() with a single column or vector though. Can any of you help me with this?

EDIT: The data below can be used to reproduce my dataset. It is not the exact values, but the format is all that is needed.

Timestamp,Author
1534334332013,user1
1534334331252,user2
1534333113577,user2
1534333112754,user3
1529160743306,user4
1528886271012,user3
1528886269171,user5
1528886261391,user5
1526477321297,user5
1526477320773,user4

Upvotes: 1

Views: 3181

Answers (2)

MrFlick
MrFlick

Reputation: 206197

The geom_bar layer will count observations for you if you just leave off the y aesthetic. For example

ggplot(msgs, aes(Author)) + geom_bar()

Tested with

msgs <- read.csv(text="Timestamp,Author
  1534334332013,user1
  1534334331252,user2
  1534333113577,user2
  1534333112754,user3
  1529160743306,user4
  1528886271012,user3
  1528886269171,user5
  1528886261391,user5
  1526477321297,user5
  1526477320773,user4")

Produces

enter image description here

Upvotes: 2

ltrd
ltrd

Reputation: 326

I produced similar example:

xxx <- rpois(1000, 10)
xtable <- table(xxx)

ggplot(data = NULL, aes(x = as.numeric(names(xtable)), y = as.numeric(xtable))) +
geom_bar(stat = "identity", width = .75)

This should produce you correct answer.

Upvotes: 2

Related Questions