CA SM - Correspondence Analysis - Show Map & Eigen Values

(3/21/13)

PURPOSE

Print selected factor map and eigenvalues.
See: Classification and Clustering Tutorial for further info.   Example.

SEE ALSO

CA S [Correspondence Analysis - Solution]
CA SR [Correspondence Analysis - Reconstitute images]
CA SRD [Correspondence Analysis - Reconstitute Differential images]
CA SRI [Correspondence Analysis - Reconstitute Importance images]
CA VIS [Correspondence Analysis - Create Visual map]
CL CLA [Classification - Automatic]

USAGE

.OPERATION: CA SM

.IMAGE OR PIXEL COORDINATES (I/M): I
[Enter 'I' if you wish to have a map of images, and 'P' for a map of pixels.]

.COORDINATE FILE PREFIX: coran_may
[Enter the prefix identifying the data files created from a previous 'CA S' run: _IMC, _PIX, & _EIG. Do not include trailing underscore.]

.NO. OF HORIZONTAL PATCHES: 0
[The number of horizontal patches does not need to be specified, UNLESS this particular CORAN analysis was done on a set of correlation patch averages. In this case, a small lookup map is printed on the sheet preceding the actual map, which helps you trace the image numbers back to the patches they originated from in the image field. E.g., a set of 18 patch averages originating from 3 rows of 6 patches each would require the input: NO. OF HOR. PATCHES = 6, and this would yield the lookup map:

1 2 3 4 5 6
7 8 9 10 11 12
13 14 15 16 17 18

.TWO FACTOR NUMBERS FOR MAP (E.G: 1,5): 1,2
[Enter the number identifying the factors for which the map is to be printed. Factor 1 is the first non-trivial factor.]

.SYMBOL, ASSIGN SYMBOL, CLASS, DOC, ID (S/A/C/D/J/I): I

[Enter 'S' if you want to enter a character symbol for each active image;
Enter 'I' if you want the numbers identifying the images or pixels to be printed. (this is the standard option for a CORAN map.);

[The following additional options are for images only:
Enter 'A' if you want to assign symbols to specific ranges of ID numbers.
Enter 'C' if you want the class number of the image to appear as symbol (in this case, a cluster file from a previous run of 'CL CLA must be available).
Enter 'D' if you want to supply a specific symbol for each image. (in this case, a document file must be supplied that specifies the symbol lookup)].

If 'S', SPIDER will ask:

.1 CHAR. SYMBOL FOR ACTIVE IMAGE/PIXEL: #
For all options, SPIDER will ask:

.PREPARE POSTSCRIPT FILE FOR MAP? (Y/N): Y
[Option 'Y' prepares Postscript output containing the map. Option 'N' lists map in RESULTS file only.]

If the answer is "N" SPIDER will ask:

.NUMBER OF PAGES (1,2,3) OR <CR>=1: 2
[Specify number of pages for the map.]

.NUMBER OF LINES OR <CR> FOR DEFAULT: <CR>
[Default: horizontal space, known from the number of pages, is used to determine the number of lines. Enter a number if you want to override the default.]

For all answers, SPIDER will ask the following two questions:

.NUMBER OF SD OR <CR>=2.3: <CR>
[The map is truncated for efficient visualization of the data cloud. Normal cutoff is set at 2.3 x standard deviation. Outliers outside the map are listed along with their coordinates, and printed exactly on the border.]

.1=FLIP #1/ 2=FLIP #2/ 3=FLIP 1+2/ <CR>=NO FLIP:<CR>
[Flips axis of map.]

In case the 'C' option was selected SPIDER will ask:

.CLUSTER FILE: CLU002
[Enter the name of the cluster file from a previous run of 'CL CLA

In case the 'A' option was selected, SPIDER will ask:

.SYMBOL: X
[Enter symbol]
.Until ID #: 320
[Enter upper limit up to which the symbol should be used]
.SYMBOL: y

.Until ID #: 320

[The questions continue until 0 or <RETURN> is entered for ID #.]

In case the 'D' option was selected, SPIDER will ask for name of document file containing the symbol lookup:

.LOOKUP FILE: DOC001
[the document file should contain, the original image ID number (=Key), and the standard Fortran code number of the ASCII character to be used as symbol (=Register 1)].

If a Postscript map was requested SPIDER will ask following questions. Usually the default <CR> answer is OK unless you want 'pretty' axes or labels:

.POSTSCRIPT OUTPUT FILE: posmap
[Default extension is .ps.]

.TEXT SIZE FOR AXIS AND DATA (USE <CR> FOR DEFAULT = 10): 12, 12
[Allows control over text size. A size of 10 is a reasonable default. A negative answer will skip the following questions and use appropriate default values instead.]

.X AXIS OFFSET: 0

Bounds for X axis of plot listed.

.NEW LOWER, UPPER AXIS BOUNDS OR <CR>: <CR>
[Can change y axis bounds or keep default by entering <CR>.]

X axis label units listed.

.NEW AXIS LABEL UNIT AND TICKS / LABEL OR <CR>: <CR>
[Can change X axis label unit or keep default]

.LABEL NO. TO EDIT. USE 0 FOR LABEL: 1
-1 TO RELIST ALL LABELS, OR <CR> TO CONTINUE: <CR>

Bounds for Y axis of plot listed.

.NEW LOWER, UPPER AXIS BOUNDS OR <CR>: <CR>

.Y AXIS OFFSET: 0

Y axis label units listed.

.NEW AXIS LABEL UNIT AND TICKS / LABEL OR <CR>: <CR>

.LABEL NO. TO EDIT. USE 0 FOR LABEL: 1
-1 TO RELIST ALL LABELS, OR <CR> TO CONTINUE: <CR>

NOTES

  1. A list of eigenvalues, their percentages, and cumulative percentages will be written in the SPIDER results file.

  2. See: Classification and Clustering Summary and Classification and Clustering Tutorial for further info.

SUBROUTINES

SGRAF, HISMAP, HISMAP4, EPUR4, POSTRT

CALLER

UTIL1

© Copyright Notice /       Enquiries: spider@wadsworth.org