Databases
StackOverflow2013
Postsdbo
PO
 

Note that there are some explanatory texts on larger screens.

plurals
  1. PO
    primarykey
    data
    text
    <h1>Alternative solution : Changing the variance factor</h1> <p>Changing the method of the variance of the random effect, seems to fix the problem.</p> <p>e.g:</p> <pre><code>mod.aic &lt;- coxph(Surv(time, event) ~ x + frailty.gamma(cluster, eps=1e-10, method="aic", sparse=0), outer.max=1000, iter.max=10000, data=dat) plot(survfit(mod.aic), col=4) </code></pre> <p><img src="https://i.stack.imgur.com/zYYIQ.png" alt="enter image description here"></p> <h1>ddd hoc solution: works if we remove one cluster</h1> <p>Maybe this don't answer exactly your question , but when I remove any cluster e.g:</p> <pre><code>par(mfrow=c(2,3)) res &lt;- sapply( 1:6 , function(x) { mod &lt;- coxph(Surv(time, event) ~ x + frailty.gamma(cluster, eps=1e-10, method="em", sparse=0), outer.max=1000, iter.max=10000, data=subset(dat,cluster != x) ) plot(survfit(mod), col=4,main= paste ('cluster', x, 'is removed')) legend(10,1,mod$iter) }) </code></pre> <p>the coxph converge and I have the same result for all samples.</p> <p><img src="https://i.stack.imgur.com/XYLfF.png" alt="enter image description here"></p> <p>I don't have enough information about your data for further analysis but I Tried to do some comparison between different clusters.</p> <pre><code>library(ggplot2 qplot(data = dat, x=time , y = x , facets= event~cluster) </code></pre> <p><img src="https://i.stack.imgur.com/cP4lp.png" alt="enter image description here"></p> <p>I notice that 3 groups : </p> <ol> <li>clusters 1,3,5 : events uniformally distributed </li> <li>clusters 2 ,4 : events just for small times. </li> <li>cluster 6 : amazing one ( only event 1)</li> </ol>
    singulars
    1. This table or related slice is empty.
    1. POFitting a gamma frailty model in R (coxph)
      singulars
      1. PTQuestion
    1. PTAnswer
    1. USagstudy
    1. USagstudy
    plurals
    1. This table or related slice is empty.
    1. This table or related slice is empty.
    1. This table or related slice is empty.
    1. This table or related slice is empty.
    1. VO
      singulars
      1. PO
      1. This table or related slice is empty.
      1. VTUpMod
    2. VO
      singulars
      1. PO
      1. This table or related slice is empty.
      1. VTUpMod
    3. VO
      singulars
      1. PO
      1. This table or related slice is empty.
      1. VTUpMod
    1. COThanks for pointing that out. "method=aic" is however not standard and I have no idea how it performs compared to "method=em" which has been proven to give the same estimates as the EM algorithm...
      singulars
      1. PO
      1. USMarco
    2. COSo no way to use aic to select the variance even with more explanation?
      singulars
      1. PO
      1. USagstudy
    3. COSorry, I do not get the question... more explanation on what? By the way, +1 for your answer anyway
      singulars
      1. PO
      1. USMarco
 

Querying!

 
Guidance
A row detail

Detail views are divided into sections. All the information in the data section comes from columns in the selected row. The other sections display data from other, related rows.

Related data can be related in a to-one or a to-many fashion. Captions of data related in a to-many fashion link to a list view showing a filtered view of the table.

Try moving around until you find a non-empty to-many entry and click on the label to get to one. You can move back to the root by clicking on the database name in the header.

SQuiL has stopped working due to an internal error.

If you are curious you may find further information in the browser console, which is accessible through the devtools (F12).

Reload