Date:         Wed, 25 Jun 2008 10:41:30 -0400
Reply-To:     Ya Huang <ya.huang@AMYLIN.COM>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Ya Huang <ya.huang@AMYLIN.COM>
Subject:      Re: Plotting measurements before and after onset of disease
Comments: To: erikjoas@GMAIL.COM

We usually call it spaghetti plot, the overlay is done by using
the third var in gplot:

data xx;
input id conc time;
cards;
1 20 -10
1 23 -5
1 45 0
1 78 5
1 70 10
2 30 -10
2 33 -5
2 55 0
2 98 5
2 80 10
;

symbol i=j repeat=100;
proc gplot data=xx;
plot conc*time=id;
run;
quit;

On Wed, 25 Jun 2008 07:00:19 -0700, E Joas <erikjoas@GMAIL.COM> wrote:

>I would like to see how levels of a certain proteins change before and
>after the diagnosis of a certain form of cancer.
>
>The data is a large population survey so not all people are affected
>by this type of cancer. I have measurements at 6 fixed times (the same
>years for all individuals) of the protein and also a variable of the
>year the disease is diagnosed.
>
>My idea is start by constructing a plot where, in the middle of the x-
>axis there is the point where the disease is diagnosed and then on
>each sides are the measurements before and after the onset of the
>disease. This is just for me to get a general plot of how these
>measurements are correlated and then proceed with further analysis.
>
>The problem is that I don?t really know how to compute the plot. Could
>you for instance create a bunch of overlay plots for each individual
>where I can specify for each individual that the onset of the disease
>should land in the middle? and how do I do this?
>
>
>Amount of x in the bloodstream
>|
>|
>|
>|
>|
>|
>|
>|
>|
>|
>|
>|
>_______-10______-5_______onset of disease____5_____10______> years
>after disease
>
>
>
>All help is welcomed and appreciated.
>Erik Joas