The program calculates Linear Discriminant Analysis (LDA) parameters using the train data separated onto two classes. The Linear Discriminant Analysis is commonly used techniques for data classification. This method maximizes the ratio of between-class variance to the within-class variance in dataset thereby guaranteeing maximal separability. The approach calculates Linear Discriminant Function (LDF) which coefficients are chosen so that they result in the best separation among the groups for train data set. Variables for the classification should be specified by the user; classes for the data should be specified in the ClassVar variable.

The LDF can be applied in the LDAClass procedure to separate any data into two groups depending on whether the value of LDF is greater or less than 0.

This program use statistical functions from "R" free software environment for statistical computing and graphics (http://www.r-project.org).

This program requires the R-package to be installed on your computer.

Input file should contain table of numerical data: lines for observations (cases) columns should be separated by tabulation or user-defines symbol (; , etc); for example, if comma (,) separator is used, the file format is the same as the CSV (comma separated values) format. No missing data allowed.

Example of output data:


LDA Statistics for class variable ClassVar:
NCASES=50; NCLASS0=20; NCLASS1=30
Var	Mean0	   Mean1	LDF
Feat1	9.3970	 -10.6047	-5.0675
Feat2	3.2846	 -3.1118	-0.6547
Feat3	1.6290	 -0.9977	1.0895
Feat4	-2.9638	 2.7626		1.1494
Feat5	-10.0696 10.0585	5.8385
B0	  	*		   *		-3.1990

First line is the header. Second line is the sample description: NCASES - number of cases total; NCLASS0 - number of class 0 cases; NCLASS1 - number of class 1 cases. Next line is output data description: Var - name of variable; Mean0 - mean for class 0; mMean1 - mean for class 1; LDF - coefficient of the linear discriminant function for the variable and b0 coefficient (B0).

Example of input data file format:


ItemName   Feat1	Feat2		Feat3		Feat4		Feat5	   ClassVar
Item1	-11.761101	-5.295846	-2.491684	4.151158	9.777093	1
Item2	-11.425886	-6.753716	0.136692	5.161748	13.618702	1
Item3	-7.069796	0.545457	0.097140	0.678579	10.302988	1
Item4	-13.480880	-3.867702	0.119297	2.333842	10.992096	1
Item5	-9.707938	-2.597949	-2.329997	2.928526	8.441053	1
Item6	-10.013794	-2.165258	-3.169195	2.625904	10.611103	1
Item7	-9.057161	-4.766594	1.691733	1.655782	7.046236	1
Item8	-8.562761	-1.272652	-3.990204	2.286294	12.768212	1
Item9	-12.724631	-4.710623	-2.114719	2.812189	6.434645	1
Item10	-9.593738	-5.478652	-1.799524	4.306497	9.514756	1
Item11	-7.699759	-1.546648	-0.423322	4.889767	9.228675	1
Item12	-13.158116	-2.891354	0.595935	2.264199	12.004761	1
Item13	-10.509598	-3.414075	-1.962310	1.263863	10.199896	1
Item14	-6.547624	-3.594928	-2.117222	5.168950	10.838221	1
Item15	-12.375988	-3.130436	-2.169164	1.537614	11.112888	1
Item16	-12.953032	-2.805048	0.085116	3.303354	7.405194	1
Item17	-11.370708	-2.848384	-0.848201	3.885525	10.569231	1
Item18	-13.117222	-7.025575	1.406507	7.069338	12.230415	1
Item19	-11.573168	0.288003	-2.826167	4.397137	10.851711	1
Item20	-7.993835	-1.204352	-1.924345	0.829829	10.314768	1
Item21	-9.225135	-2.512925	-1.608051	1.420301	9.766411	1
Item22	-8.402783	-0.890500	3.189703	3.754479	7.481063	1
Item23	-9.888180	-3.345775	1.965667	2.906369	11.488815	1
Item24	-11.686270	-5.389477	2.556932	1.661153	9.717826	1
Item25	-12.599567	-0.266091	-3.936308	0.751762	10.405225	1
Item26	-11.365093	-1.919706	-0.458052	1.861843	9.521104	1
Item27	-11.027619	-2.944884	-2.792962	4.144322	7.958556	1
Item28	-11.795160	-6.769646	0.908383	1.005066	11.240333	1
Item29	-13.629933	0.674184	-3.386853	-0.095859	10.490432	1
Item30	-7.823298	-5.452589	-2.336894	1.919889	9.421125	1
Item31	6.360118	6.794549	4.168188	-4.492538	-12.297555	0
Item32	8.774682	0.492721	1.587909	-5.486587	-12.361278	0
Item33	7.768181	3.989776	3.289377	-0.895444	-13.067171	0
Item34	8.581133	2.922361	3.952544	-4.450362	-6.787133	0
Item35	6.176519	4.526292	-2.771599	-3.477187	-7.316202	0
Item36	11.539781	-0.892880	2.868221	-1.456557	-11.008881	0
Item37	11.743034	3.527726	-0.635792	-2.067965	-7.151524	0
Item38	10.527299	1.460768	0.862300	-1.967742	-8.819727	0
Item39	8.148808	5.157964	-0.916135	-1.551958	-9.467513	0
Item40	9.241432	1.483108	-0.981933	1.046571	-8.504166	0
Item41	9.444197	4.963927	1.127201	-0.523484	-9.102817	0
Item42	11.545396	4.604968	4.818171	-5.046815	-13.494675	0
Item43	11.890988	1.220710	-2.069796	-2.942747	-8.996673	0
Item44	11.810480	2.031465	2.987976	-5.699606	-10.026246	0
Item45	10.806543	5.275155	4.969420	-2.792596	-11.345561	0
Item46	10.261177	3.586077	3.340220	-3.339244	-7.795038	0
Item47	8.407544	2.887997	3.104312	-3.734519	-9.758477	0
Item48	7.317484	5.553850	1.618000	-2.525315	-13.613147	0
Item49	10.654500	2.579577	1.922452	-3.765160	-10.414136	0
Item50	6.940641	3.525834	-0.660756	-4.105869	-10.064455	0