A propos du package SAISIR -- Site web Officiel
Exemple de fonctions présentes dans SAISIR
A propos de SAISIR
Graphics : biplot, curves, confidence ellipse, barycenter representation, dendrogram ...
Factorial methods : PCA, correspondence analysis, factorial distance analysis ...
Discrimination : linear, factorial, quadratic , PLS, stepwise ...
Regression : PLS1, PLS2, PCR, ridge, latent root , stepwise ...
Multiway table : statis, ACCPS, Multiple factor analysis, canonical analysis ...
SAISIR means "Statistics applied to the Interpretation of Spectra in the Infrared". The French word "saisir" means "to grasp" or (metaphorically) "to understand".
SAISIR has been initially designed by a spectroscopist who was not very satisfied of the limitation of many specific environment for processing spectra. This environment can now be used for most of the chemometric work or for data mining whatever is the kind of data. It gives a meeting place in which every kind of data can interact with each others.
It contains the basic methods of chemometrics (such as PLS or Principal component analysis) but also many other chemometric methods. SAISIR. Contrary to what happens in most chemometric environment, it is very easy to get in the package from Matlab, or to get out in order to benefit of the complete Matlab environment. The more skilled user can easily find in SAISIR sample of code that can be reused or modified for in-house applications.
In its current form, SAISIR includes about 200 functions. In the loadable package, the user will find typical script for the more common situations (data processing, regression, discrimination, application of factorial methods, data pre-processing). The package also contains a powerpoint tutorial with exercises and worked examples making it possible to learn SAISIR (and perhaps Matlab) within a few days.
Chimiométrie.fr
Vous êtes le ième visiteurs
Dernière mise à jour : 07/10/2016