Skip to content

Plotting LDA results

4 messages · Meffy, Uwe Ligges, David L Carlson

#
Dear Users!
I think I still have some problems in understanding LDA and the methods of
plotting the results.
The case is the following: I'm having a dataset containing two classes where
each datapoint has 19 dimensions. Training with lda(...) works fine, and I'm
getting 19 LD coefficients. So far so good. Now I want to visualize the
result, and here is where my simple knowledge ends. What I simply want to do
is doing a scatterplot on two dimensions and then plotting the projected
hyperplane in order to see the dividing line. The scatterplot I'm doing with 

plot(class1[1:100,1], class1[1:100,2])
points(class2[1:100,1], class2[1:100,2])

but now I face with the problem of determining the intercept for abline(...)
(the slope should be -LD[1]/LD[2] I think...).

What I really don't understand is the result of plot.lda(...). As a result
it gives two histograms, each one for one class. So which distributions are
shown there? And why does it only plot a scatterplot when I have three
classes? 

Thank you very much in advance!
Greetings, Matthias








--
View this message in context: http://r.789695.n4.nabble.com/Plotting-LDA-results-tp4637766.html
Sent from the R help mailing list archive at Nabble.com.
#
On 25.07.2012 13:52, Meffy wrote:
The scatterplot shows a projection on the first two linear discriminant 
components, the histogram is the projection on the first component only.
>
If you have two classes, there will only be one component. No chance two 
invent a second dimension.

Uwe Ligges
#
Thank you!
Still not clear...I can plot two of my data dimensions against each other
where I see two separated clouds. So it must be possible to determine the
coefficients of the dividing line from the model I get from lda(...), or am
I completely wrong now...?



--
View this message in context: http://r.789695.n4.nabble.com/Plotting-LDA-results-tp4637766p4637780.html
Sent from the R help mailing list archive at Nabble.com.
1 day later
#
You asked why you could not get two discriminant functions and that question
was answered. The number of discriminant functions is one less than the
number of groups (assuming you have more variables than groups). Now you are
asking a different question. How to plot the discriminant boundary between
the groups in a plot of the original variables. Time to provide "provide
commented, minimal, self-contained, reproducible code" [including data] as
called for at the bottom of your message.

----------------------------------------------
David L Carlson
Associate Professor of Anthropology
Texas A&M University
College Station, TX 77843-4352