s-news
[Top] [All Lists]

Predicting Hazard Rates with censorReg

To: <s-news@lists.biostat.wustl.edu>
Subject: Predicting Hazard Rates with censorReg
From: "Andrew Scott" <andrew.scott@icfrith.com.au>
Date: Fri, 7 Apr 2006 17:38:17 +1000
Cc: <enquiries@solutionmetrics.com.au>
Thread-index: AcZaFjxyxOifEh9eTOeUyBwMIpBI0A==
Thread-topic: [S] Predicting Hazard Rates with censorReg

Dear S-News Readers,

 

I’ve got a model based on censorReg using the following:

 

SurvCurve.2<-censorReg(censor(startTime,endTime,status)~1,data="">

 

Now I want to predict proportions failed with the following:

 

y<-predict(SurvCurve.2,type="probability",q=endTimeSum)

 

where my predicted durations are as follows:

 

> endTimeSum

[1] 15 24 27 36 48 60 72 96

 

This gives me the following predictions:

 

> y

$"(intercept)":

     Estimate    Std.Err   95% LCL    95% UCL

15 0.03122837 0.03736268 0.0290874 0.03352149

24 0.17852421         NA        NA         NA

27 0.26701616 0.02628998 0.2570531 0.27722129

36 0.61280592 0.08752239 0.5714039 0.65263977

48 0.94487842 0.27288420 0.9094281 0.96695744

60 0.99898308 0.78580469 0.9952734 0.99978185

72 0.99999915 1.87726321 0.9999665 0.99999998

96 1.00000000 0.00200296 1.0000000 1.00000000

 

OK so far.

 

Now to my questions:

 

(1)     I want to pick off the vector of estimated failure rates above (i.e. 0.0312, 0.1785,….., 1.0000). How can I do this?

(2)     Then I want to multiple (1) by an equivalent vector of units at risk. The number of units by duration is as follows:

 

> x<-table(endTime)

> x

            15         24         27         36         48         60         72         96

            1          473        7         26155   2218     9285     35         13

 

Thus the final calculation will be the multiplication of two vectors (0.0312, 0.1785,….., 1.0000) and (1, 473, 7, ….., 13).

 

Any help would be appreciated.

 

Andrew Scott
Manager - Strategic Analysis
Tel: 02 8853 9117
Mob: 0411 555 486
Fax: 02 8853 9517
andrew.scott@icfrith.com.au

IC Frith & Associates (Australia’s Warranty Provider)
Suite 1, Building B, 34-46 Brookhollow Ave, Baulkham Hills NSW 2153
Private Mail Bag 14, Castle Hill 1765
Phone: 02 8853 9100
Fax: 02 9634 2396
www.icfrith.com.au


The information contained in this email and any attached files is strictly confidential and may contain legally privileged information. This email should be read by the intended addressee only. If the recipient of this message is not the intended addressee, please advise ICF (Australia) Pty Ltd on +61 2 8853 9100 or icfteam@icfrith.com.au and promptly delete this email and any attachments. The intended recipient of this email may only use, reproduce, disclose or distribute the information contained in this email and any attached files with ICF (Australia) Pty Ltd's permission. If you are not the intended addressee, you are strictly prohibited from using, reproducing, disclosing or distributing the information contained in this email and any attached files. Although ICF (Australia) Pty Ltd believes this e-mail and any attachments to be free of any virus or other defect which may effect your computer, it is the responsibility of the recipient to ensure that it is virus free and ICF (Australia) Pty Ltd does not accept any responsibility for any loss or damage arising in any way from its use. We maintain a database of clients' contact details. If you do not wish to receive electronic messages from us in future please email icfteam@icfrith.com.au. Finally, you should be aware ICF (Australia) Pty Ltd reserves the right and intends to intercept and monitor incoming and outgoing e-mail correspondence, so you should not expect any e-mail communication to be private in nature. ICF (Australia) Pty Ltd ABN 75 115 419 853 AFSL No:291260

 

<Prev in Thread] Current Thread [Next in Thread>
  • Predicting Hazard Rates with censorReg, Andrew Scott <=