Dear list I'd like to have your opinion about my case study.
I'm analizing a dataset of 9 experiments and 15 variables with the aim to
highlight the variables that can majorly explain the variance between the
experiments.
This is an example with only 3 rows and 5 variables
var1 var2 var3 var4 var5 sample5 0,067
0,005 0,008 0,100 0,005 sample6 0,069 0,001 0,011 0,084 0,005 sample7 -7
-5 -1 34 4
My problem is that in some experiments (like in sample7) the measures
related to my variables are measured as delta values (initial condition -
final condition). In the other cases the variables are measured considering
only the absolute values at my final condition.
After PCA the model looks like strongly influenced by this difference
(even if my data are centered to 0 and scaled to 1) because in the score
plot I see with the first PC mainly the separation between experiments with
positive and negative values and the second PC is not able to give to me
further informations .
In your opinion is there a way to compare these experiments measured in
this different way?
Alternatively do you think that the Dual Multiple Factor Analysis available
with the package FactorMineR could be a better way to analyze these data?
Thank you for any suggestion
Guido
--
Guido Leoni
National Research Institute on Food and Nutrition
(I.N.R.A.N.)
via Ardeatina 546
00178 Rome
Italy
tel + 39 06 51 49 41 (operator)
+ 39 06 51 49 4498 (direct)
[[alternative HTML version deleted]]
______________________________________________
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
