From: Jian Cao
Subject: RE: a question about JMP platform of fit-y-by-x and Multivariate
Date: April 24, 2014 at 10:50:41 AM EDT
To: Lorry Li , "[JMP] SE"
Lorry,
If you create a discrete but numerical variable (SEX2) and use value labels, then you will be able to compute correlations between sex and all other variables.
use
Jian
From: Lorry Li
Sent: Thursday, April 24, 2014 10:20 AM
To: SE
Subject: a question about JMP platform of fit-y-by-x and Multivariate
Dear SEs,
A JMP user asked me a question about JMP platform of fit-y-by-x and Multivariate, I try to re-interpret it as following:
1. for numeric data, JMP’s Multivariate platform can generate a Scatterplot Matrix to show all the two-way correlations, but it does not work for discrete data;
2. with JMP’s platform of Fit-Y-by-X, JMP can generate the following graphs to show the two-way correlation between two variables of which at least one variable is discrete.
Mosaic Plot
the JMP user is asking whether it is achievable to put all the graphs (plots) together.
In order to make myself clear, I PSed a picture as following to show what it is expected to be:
Does anybody know whether it is realizable?
Thanks a lot!
Lorry Li | 李红波
JMP Greater China 敏捷分析 成就无限
M:86 18616575537
T: 86 21 61633031
F: 86 21 61638589
Web: www.jmp.com/china
Forum: www.jmpforum.net
Addr.:1504, AZIA Center, No.1233 Lujiazui Ring Road, Shanghai PRC 200120
上海市浦东新区陆家嘴环路1233号汇亚大厦1504室（200120）
