Process Help |
Process Name |
Menu Path |
Link to Command Table |
DISCAN |
Sample Analysis ribbon | Geochemical Processes | Multiple Discriminant Analysis |
Introduction
The calculation of discriminant functions (Discriminant analysis) will establish a set of rules using geological control groups or training sets which can then be applied to the classification of unknown samples using the DISCLA process.
How to use
As a check, the rules as defined by the discriminant functions and group centroids can then be used to reclassify the training sets to see how many mismatches occur.
The linear discriminant function can be considered as a planer separating boundary between the different groups as classified by their input variables (fields), eg. CU, PB and ZN. The group centroids are the fixed positions in space locating the centres of each group cluster as defined by the variables.
File Handling
The input file (&IN) must contain a sample identifier field (@SAMPID) and a group identifier field (@GROUPID). Data from the different control groups must have a unique identifier common to all samples within the group. All the different control groups must be put into one file using the APPEND process before using DISCAN. The group identifier will be representative of a geological control group, eg. to identifier them as granite, mineralised zone, background, lithology A or lithology B etc..
There are optional output files for the calculated discriminant functions (&FUNCTS), group centroids (&CENTRDS) and discriminant scores *&(SCORES). If the multiple discriminant functions and group centroids are required to classify unknown data in the DISCLA process then the relevant output files must be named. If an output file of the re-classified control data is required then the file &SCORES must be used.
If the user wishes to plot maps of the output scores then the scores file can be joined to the original input file using the JOIN process and defining SAMPID as the keyfield.
Files, Fields and Parameters
Input Files
Name |
Description |
I/O Status |
Required |
Type |
IN |
Input file. Must contain group and sample identity fields. A maximum of ten groups are evaluated. The file must be sorted on the *GROUPID field. |
Input |
Yes |
Undefined |
Output Files
Name |
I/O Status |
Required |
Type |
Description |
CENTRDS |
Output |
No |
Undefined |
Optional output file containing group centroids identified by the field named in GROUPID. |
FUNCTS |
Output |
No |
Undefined |
Optional output file containing discriminant functions. |
SCORES |
Output |
No |
Undefined |
Optional output file containing the calculated scores for the individual samples and a misclassification field. |
Fields
Name |
Description |
Source |
Required |
Type |
Default |
GROUPID |
Compulsory group identifier field contained in input file IN. |
IN |
Yes |
Alphanumeric |
Undefined |
SAMPID |
Compulsory sample identifier field contained in input file IN. |
IN |
Yes |
Any |
Undefined |
F1 |
First field to be used. No fields specified means all. |
IN |
No |
Numeric |
Undefined |
F2 |
Second field to be used. |
IN |
No |
Numeric |
Undefined |
F3 |
Third field to be used. |
IN |
No |
Numeric |
Undefined |
F4 |
Fourth field to be used. |
IN |
No |
Numeric |
Undefined |
F5 |
Fifth field to be used. |
IN |
No |
Numeric |
Undefined |
F6 |
Sixth field to be used. |
IN |
No |
Numeric |
Undefined |
F7 |
Seventh field to be used. |
IN |
No |
Numeric |
Undefined |
F8 |
Eighth field to be used. |
IN |
No |
Numeric |
Undefined |
F9 |
Ninth field to be used. |
IN |
No |
Numeric |
Undefined |
F10 |
Tenth field to be used. |
IN |
No |
Numeric |
Undefined |
Parameters
Name |
Description |
Required |
Default |
Range |
Values |
||||||
RESUM |
|
No |
0 |
0,1 |
0,1 |
||||||
PRIMAT |
|
No |
0 |
0,1 |
0,1 |
||||||
PRISCO |
|
No |
0 |
0,1 |
0,1 |
Notes
There is a maximum limit of 10 control groups and 45 variables. All individual control groups must also be appended into one file prior to analysis.
Example
!DISCAN
|
&IN(GG), &CENTRDS(GGCENT), &FUNCTS(GGFUNCTS), @GROUPID='GROUPID', @SAMPID= 'ID', @RESUM=1, @PRIMAT=1 |
Error and Warning Messages
Message |
Description |
Solution |
*** Warning *** GROUPID field has nn words. Only the first 3 will be used. |
|
|
|
|
|
*** Warning *** SAMPID field has nn words. Only the first 3 will be used |
|
|
|
|
|
*** Error *** GROUPID field (fieldname) is not in input file. |
|
|
|
|
|
*** Error *** No numeric fields on file (filename) |
|
|
|
|
|
*** Error *** SAMPID field (fieldname) is not in input file. |
|
|
|
|
|
*** Error *** in reading data from
file (filename) |
|
|
|
|
|
*** Error *** There is only one group in file
(filename) |
|
|
|
|
|
*** Error *** There is only one sample in
each group |
|
|
|
|
|
*** Error *** Compulsory file, field or parameter
missing |
|
|
|
|
|