Re: Fwd: vista question
Neal Lathia
Many thanks for your email - as you suggested, I've cc'ed my reply to
the group.
As I said before, I'm trying to use Vista 6.4 (with the plugins) for
correspondence analysis. After my previous problems, I decided to go
through the example in Chapter 11 of the manual.
I copied the data from the chapter by doing File > New Edit, and saving
what I put in as a .lsp (attached - doctorates_input.lsp). However, when
I go to the Analyze menu, the Correspondece Analysis option cannot be
selected.
I have also attached the vista file (doctorates_vista.lsp). Any thoughts
as to what I've done wrong?
Many thanks,
Neal
---
http://www.cs.ucl.ac.uk/staff/n.lathia
Pedro Valero wrote:
> Dear Neal,
>
> I am writing you on behalf of Patty Young, the wife of Forrest. Forrest
> passed away a few months ago and since then she forwards the messages
> addressed to him to me as we collaborated with regard to ViSta in the
> last years of his life.
>
> I am not completely sure of the process you followed. How did you pass
> the data from Excel to ViSta. Did you save as .txt in Excel and then
> imported into ViSta? If you did and it worked smoothly the problem you
> may have is with the datatype. Forrest worked a lot on this issue but I
> am afraid it did not ever set it completely right. If you send me the
> ViSta file and I will fix it for you (and I will explain how to do it in
> the future).
>
> We have set a new discussion group for ViSta in the following address
>
> http://www.uv.es/visualstats/Book/DiscussionGroup.htm
>
> As we do not have much traffic yet, I would appreciate we go on this
> discussion in the group.
>
> Best,
>
> Pedro
>
>
> ================================================
> Dr. Pedro M. Valero-Mora (val...@uv.es)
> Departamento de Metodología de las Ciencias del Comportamiento
> Facultad de Psicología
> Universitat de Valencia
> C/Blasco Ibáñez, 21
> Valencia
> España
> CP 46010
> Tlf. (34) (963) 64507
> Fax(34) 963864697
> If you have read until here, check the Visual Statistics book too!
> http://www.uv.es/visualstats/Book/
>
Pedro Valero
The problem you have is related to the datatype of data. In Forrest's
terminology, correspondence analysis works with frequency table data,
i.e. the crosstabulation of two categorical data. This is controlled by
the parameter :freq in the dataset file. In your case, the file has
:freq nil. Changing it to to :freq t does the trick (see file pasted
below).
ViSta 6 did not provide any mechanism to indicate that the data were
frequency table data when importing from a .txt file. However, if I
remember correctly, ViSta 7 does. You can download ViSta 7 at the
address of the Visual Statistics book (go to
http://www.uv.es/visualstats/Book and then follow the links).
(DATA "Doctorates" :TITLE "Science Doctorates Conferred in the USA,
1960-1975" :ABOUT "There is no information about these data." :FREQ t
:ROW-LABEL NIL :COLUMN-LABEL NIL :ARRAY NIL :VARIABLES (QUOTE ("60"
"65" "70" "71" "72" "73" "74" "75")) :TYPES (QUOTE ("Numeric" "Numeric"
"Numeric" "Numeric" "Numeric" "Numeric" "Numeric" "Numeric")) :LABELS
(QUOTE ("Engineering" "Mathematics" "Physics" "Chemistry"
"EarthSciences" "Biology" "Agriculture" "Psychology" "Sociology"
"Economics" "Anthropology" "Others")) :DATA (QUOTE (794 2973 3432 3495
3475 3338 3144 2959 291 685 1222 1236 1281 1222 1196 1149 530 1046 1655
1740 1635 1590 1334 1293 1078 1444 2234 2204 2011 1849 1792 1762 253
375 511 550 580 577 570 556 1245 1963 3360 3633 3580 3636 3473 3498 414
576 803 900 855 853 830 904 772 954 1888 2116 2262 2444 2587 2749 162
239 504 583 638 599 645 680 341 538 826 791 863 907 833 867 69 82 217
240 260 324 381 385 314 502 1079 1392 1500 1609 1531 1550))
:DATASHEET-ARGUMENTS (QUOTE ((1393 -15) (0 879) 2 8)))
Neal Lathia ha escrito:
> --------------040403000502000808050504
> Content-Type: text/plain
> Content-Disposition: inline;
> filename="doctorates_vista.lsp"
>
>
> (DATA "Doctorates" :TITLE "Science Doctorates Conferred in the USA, 1960-1975" :ABOUT "There is no information about these data." :FREQ NIL :ROW-LABEL NIL :COLUMN-LABEL NIL :ARRAY NIL :VARIABLES (QUOTE ("60" "65" "70" "71" "72" "73" "74" "75")) :TYPES (QUOTE ("Numeric" "Numeric" "Numeric" "Numeric" "Numeric" "Numeric" "Numeric" "Numeric")) :LABELS (QUOTE ("Engineering" "Mathematics" "Physics" "Chemistry" "EarthSciences" "Biology" "Agriculture" "Psychology" "Sociology" "Economics" "Anthropology" "Others")) :DATA (QUOTE (794 2973 3432 3495 3475 3338 3144 2959 291 685 1222 1236 1281 1222 1196 1149 530 1046 1655 1740 1635 1590 1334 1293 1078 1444 2234 2204 2011 1849 1792 1762 253 375 511 550 580 577 570 556 1245 1963 3360 3633 3580 3636 3473 3498 414 576 803 900 855 853 830 904 772 954 1888 2116 2262 2444 2587 2749 162 239 504 583 638 599 645 680 341 538 826 791 863 907 833 867 69 82 217 240 260 324 381 385 314 502 1079 1392 1500 1609 1531 1550)) :DATASHEET-ARGUMENTS (QUOTE ((1393 -15) (0 879) 2 8)))
> --------------040403000502000808050504
> Content-Type: text/plain
> Content-Disposition: inline;
> filename="doctorates_input.lsp"
>
> (data "Doctorates"
> :title "Science Doctorates Conferred in the USA, 1960-1975"
> :variables '("60" "65" "70" "71" "72" "73" "74" "75")
> :labels '("Engineering" "Mathematics" "Physics" "Chemistry"
> "EarthSciences" "Biology" "Agriculture" "Psychology"
> "Sociology" "Economics" "Anthropology" "Others")
> :data '(
> 794 2973 3432 3495 3475 3338 3144 2959
> 291 685 1222 1236 1281 1222 1196 1149
> 530 1046 1655 1740 1635 1590 1334 1293
> 1078 1444 2234 2204 2011 1849 1792 1762
> 253 375 511 550 580 577 570 556
> 1245 1963 3360 3633 3580 3636 3473 3498
> 414 576 803 900 855 853 830 904
> 772 954 1888 2116 2262 2444 2587 2749
> 162 239 504 583 638 599 645 680
> 341 538 826 791 863 907 833 867
> 69 82 217 240 260 324 381 385
> 314 502 1079 1392 1500 1609 1531 1550))
> --------------040403000502000808050504--
Neal Lathia
Many thanks for your help. I got it to work, using Vista 7 and your
suggestions.
The only further question I had is; is it possible to get the
coordinates used in the spin and the scatter plots of each individual
variable? I looked in the model reports but was unable to find them there.
Once again, thanks.
neal
Pedro Valero
Use the create data menu item in the model menu
P
Neal Lathia ha escrito: