Generate an R lowess line graph

We can generate a lowess line on top of the Scatter diagram. The lowess line would likely show a better fit as smoothing is used to fit the line to the data.

How to do it...

We can use the same Scatter diagram as the basis and then call upon lines to add our lowess line:

# load the iris dataset
data <- read.csv("http://archive.ics.uci.edu/ml/machine-learning-databases/iris/iris.data")

#Let us also clean up the data so as to be more readable
colnames(data) <- c("sepal_length", "sepal_width", "petal_length", "petal_width", "species")

# call plot first
plot(data$sepal_length, data$petal_length)

# add the lowess line to the graph
lines(lowess(data$sepal_length, data$petal_length), col="blue")

Here is the resulting graph:

How it works...

Locally weighted scatterplot smoothing (LOWESS) is a useful mechanism when performing a regression to see a smoothed line drawn through our data. Once accomplished, you are likely to see relationships and possibly forecasts...