Plot mean, standard deviation, standard error of the mean, and confidence interval

Question

I'm using R to plot statistical graphs. I can plot simple graphs. However, I not have idea how to plot complex data.

Someone could help me put together a chart with all this information? Excluding N (count).

My dataset

$ df <- read.csv("database.csv", header=TRUE, sep=",")
$ df

# df is 'data frame'
# Gives count (N), mean, standard deviation (sd), standard error of the mean (se), 
# and confidence interval (ci)

                   method   N       mean          sd           se          ci
4                       A 100  0.3873552 0.014513971 0.0014513971 0.002879887
6                       B 100  0.3873552 0.014513971 0.0014513971 0.002879887
11                      C 100  0.3873552 0.014513971 0.0014513971 0.002879887
12                      D 100  0.3873552 0.014513971 0.0014513971 0.002879887
10                      E 100  0.3757940 0.027337627 0.0027337627 0.005424378
1                       F 100  0.3715910 0.006180728 0.0006180728 0.001226391
3                       G 100  0.3642126 0.051949010 0.0051949010 0.010307811
5                       H 100  0.3615370 0.006790384 0.0006790384 0.001347359
9                       I 100  0.3589878 0.010288660 0.0010288660 0.002041493
13                      J 100  0.3585191 0.030658287 0.0030658287 0.006083269
2                       K 100  0.3570351 0.013531711 0.0013531711 0.002684985
7                       L 100  0.3497304 0.078784858 0.0078784858 0.015632625
8                       M 100  0.2994054 0.009305584 0.0009305584 0.001846430

I need a simple chart that contains all information of the dataset.

Note: Count (N) and standard error of the mean (se) is not necessary.

For example, how can I create a bar graph and add confidence interval?

Other e.g.:

enter image description here

Have a look at http://docs.ggplot2.org/current/stat_summary.html — Metrics, Nov 05 '13 at 20:04
I need a simple chart that contains all the information given above. — Alan Valejo, Nov 05 '13 at 20:13
@AlanValejo there are a ton of different ways to plot the information. — Señor O, Nov 05 '13 at 20:19
@ScottChamberlain For example, how can I create a bar graph and add confidence interval? — Alan Valejo, Nov 05 '13 at 20:26
@AlanValejo Have you looked at other posts on SO? It's full of examples of how to do this. Here's one: http://stackoverflow.com/questions/1432867/boxplot-schmoxplot-how-to-plot-means-and-standard-errors-conditioned-by-a-facto?rq=1 — Señor O, Nov 05 '13 at 20:37

score 6 · Accepted Answer · answered Nov 05 '13 at 23:34

6

names=data$method
x = 1:13*2-1
CI.up = as.numeric(data$mean)+as.numeric(data$ci)
CI.dn = as.numeric(data$mean)-as.numeric(data$ci)
plot(data$mean~x, cex=1.5,xaxt='n',ylim=c(0.3,0.40), xlab='',ylab='lalala!', main='blahblahblah',col='blue',pch=16)
axis(1, at=x, labels=names)
arrows(x,CI.dn,x,CI.up,code=3,length=0.2,angle=90,col='red')
legend("bottomleft",paste(names,": S.E=",data$se),ncol=6,text.width=1)

output

answered Nov 05 '13 at 23:34

Tay Shin

528
4
17

Hi, @Tay Shin, can you help solve this please? http://stackoverflow.com/questions/36167800/plot-time-series-with-vertical-error-bars-by-row-in-r?noredirect=1#comment59973528_36167800 – Joke O. Mar 23 '16 at 14:49

score 2 · Answer 2 · answered Nov 05 '13 at 20:28

2

This will plot your data.

ggplot(melt(df, id.vars=c("method", "N")), aes(method, value)) + 
geom_bar(stat="identity") + facet_wrap(~variable)

answered Nov 05 '13 at 20:28

Señor O

17,049
2
45
47

Plot mean, standard deviation, standard error of the mean, and confidence interval

2 Answers2

Linked