Dear Pavan Hande,

I have very limited experience with PCAs but I found a helpful tutorial that demonstrates how to do a PCA analysis in R*:

http://www.r-bloggers.com/computing-and-visualizing-pca-in-r/

This follows on from a post about the theory underlying PCA analyses:

https://tgmstat.wordpress.com/2013/11/21/introduction-to-principal-component-analysis-pca/

I hope these are useful. 

Yours Sincerely,

Joseph Crisp

* A helpful tutorial for R can be found here: http://www.sr.bham.ac.uk/~ajrs/R/r-getting_started.html

Dear Shivaji,

Can you maybe describe your datset a bir further? How many samples are there, and what parameters have been determined? And do you already know what you will investigate, or is your analysis more explorative of nature?

Prior to the PCA analysis, you have to think about pre-processing; this will influence the outcome to a certain degree. Next, you have to decide on the number of latent variables you are going to use. Again, this will influence the results of your analysis. Finally, you will have to interpret the results. You can use bi-plots, loading plots and/or scatter plots to draw conclusions!

Dear Pavan

Yes you can apply, provided you have sufficient ground in terms of number of data points and number of variable. 

Define your objective first. Applying PCA or to say any statistical technique must be based on the objective.

Pl see the attched paper for PCA or FA, to say. May be of some help, pl. 

Chapter A note on the Factor Analysis in Forestry Sciences