Plot multipoints and a best fit line - r

Question

I want to create one plot graph with the Roundrobin and Prediction points, without colors, where the Roundrobin and Prediction type of points are different, and it has a legend. I was want to add a best fit line for the results.
I am having trouble in adding all these features into one graph that has 2 points. I am used to Gnuplot, but I don't know how to do this with R. How I do this with R?
[1] Input data
Inputdata,Roundrobin,Prediction
1,178,188
2,159,185
3,140,175
[2] Script to generate data
no_faults_data <- read.csv("testresults.csv", header=TRUE, sep=",")
# Graph 1
plot(no_faults_data$Inputdata, no_faults_data$Roundrobin,ylim = range(c(no_faults_data$Roundrobin,no_faults_data$Prediction)),xlab="Input data size (MB)", ylab="Makespan (seconds)")
points(no_faults_data$Inputdata, no_faults_data$Prediction)
abline(no_faults_data$Inputdata, no_faults_data$Roundrobin, untf = FALSE, \dots)
abline(no_faults_data$Inputdata, no_faults_data$Prediction, untf = FALSE, \dots)
legend("top", notitle, c("Round-robin","Prediction"), fill=terrain.colors(2), horiz=TRUE)

Answer 1

In base R you will have to create a fitted model first:
robin <- lm(Roundrobin ~ Inputdata, data = no_faults_data)
pred <- lm(Prediction ~ Inputdata, data = no_faults_data)
plot(no_faults_data$Inputdata, no_faults_data$Roundrobin,
ylim = range(c(no_faults_data$Roundrobin,no_faults_data$Prediction)),
xlab = "Input data size (MB)", ylab = "Makespan (seconds)",
col = "green", pch = 19, cex = 1.5)
points(no_faults_data$Inputdata, no_faults_data$Prediction, pch = 22, cex = 1.5)
abline(robin, lty = 1)
abline(pred, lty = 5)
legend(1.1, 155, legend = c("Round-robin","Prediction"), pch = c(19,22), col = c("green","black"),
bty = "n", cex = 1.2)
which gives:
For further customization of the base R plot, see ?par and ?legend.
With ggplot2 you will need to reshape your data into long format:
library(reshape2)
library(ggplot2)
ggplot(melt(no_faults_data, id="Inputdata"),
aes(x=Inputdata, y=value, shape=variable, color=variable)) +
geom_point(size=4) +
geom_smooth(method = "lm", se = FALSE) +
theme_minimal()
which gives:
Used data:
no_faults_data <- read.csv(text="Inputdata,Roundrobin,Prediction
1,178,188
2,159,185
3,140,175", header=TRUE)

Answer 2

You should look into the ggplot2 package for plotting. Maybe not needed for the 3 points data you provided but it makes much nicer plots than the default.
df <- data.frame("Inputdata" = c(1,2,3,1,2,3), "score" = c(178,159,140,188,185,175), "scoreType" = c(rep("Roundrobin",3), rep("Prediction",3)))
p <- ggplot(data=df, aes(x=Inputdata, y=score, group=scoreType, shape = scoreType)) + geom_point(size=5)
p <- p + ggtitle("My Title")
p+stat_smooth(method="lm",se = FALSE)
Here you group by the type of score and let GG plot make the legend for you. stat_smooth is using lm here.