Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down on April 23, and its replacement, is already up and running.

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: RE; Summarise similar curves

From   Janet Hill <>
To   "" <>
Subject   Re: st: RE; Summarise similar curves
Date   Wed, 24 Apr 2013 11:56:13 +0100 (BST)


Many thanks, as usual. I will investigate the interpolations.


----- Original Message -----
From: Nick Cox <>
To: "" <>
Sent: Wednesday, 24 April 2013, 11:48
Subject: Re: st: RE; Summarise similar curves

I can't judge from your description how well a restricted cubic spline
will work. I am happy to regard -rcspline- (SSC) as a useful command
for what it does.

But the gaps in your data do not prohibit much if it's reasonable to
interpolate to a regular grid.

Check out

-ipolate- (official)
-cipolate- (SSC)
-csipolate- (SSC)
-nnipolate- (SSC)
-pchipolate- (SSC)

All these commands support applications to panel data.

A loop

foreach m in ci csi nni pchi {
    ssc desc `m'polate

will bring up capsule descriptions of the unofficial commands.


On 24 April 2013 11:34, Janet Hill <> wrote:
> I have a dataset which should have consisted of a patient characteristic measured at 6 week intervals over a two year period. There are 400 patients in three groups and I intended to graphically summarise the important features of the groups using the principal component method described by Jones,MC and Rice,JA, 'Displaying the important features of large collections of similar curves, American Statistician,46(1992) 140-145'. Unfortunately the patients did not always turn up on the specified date so I do not have an equispaced dated grid thus precluding the use of this method.
> I am now considering using a restricted cubic spline, (rcspline from to show a smoothed curve for each group.
> My questions are a) is this a reasonable approach or b) is there anything else in the Stata armamentarium which is more suitable. I am using Stata 12.

*   For searches and help try:

*   For searches and help try:

© Copyright 1996–2016 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index