tag:blogger.com,1999:blog-8499895524521663926.post8396443160622467041..comments2024-03-27T07:13:39.236-04:00Comments on Phylogenetic Tools for Comparative Biology: More on using threshBayes to measure the correlation between two binary traits in phytoolsLiam Revellhttp://www.blogger.com/profile/04314686830842384151noreply@blogger.comBlogger3125tag:blogger.com,1999:blog-8499895524521663926.post-45595538359519385472020-08-24T13:47:45.651-04:002020-08-24T13:47:45.651-04:00Thank you for your responce.
Yes, it's weird....Thank you for your responce. <br />Yes, it's weird. I have check the code with a different tree and different data (worked fine, with more "normal" values), but in the case of the dataset where I have 0, both the continous and discreat date give the same result. I don't have any missing values. The code also gives an error if there are any NAs. Kaszczakhttps://www.blogger.com/profile/08932069916668207584noreply@blogger.comtag:blogger.com,1999:blog-8499895524521663926.post-41197233131101393792020-08-24T11:26:57.562-04:002020-08-24T11:26:57.562-04:00Weird. I have seen that error when I fail to speci...Weird. I have seen that error when I fail to specify the 'types' of each character correctly (for instance, if one of the characters is a continuous character but I put types=c("disc","disc")), or, likewise, if I have missing data values in my input data. Liam Revellhttps://www.blogger.com/profile/04314686830842384151noreply@blogger.comtag:blogger.com,1999:blog-8499895524521663926.post-47348858975396777982020-08-23T09:07:06.135-04:002020-08-23T09:07:06.135-04:00Hi Liam,
thank you for the code and very helpful ...Hi Liam, <br />thank you for the code and very helpful examples. The code itself seems to be not very complicated (for my data:w<-threshBayes(tree,data,types=c("disc","disc"),ngen=100000,control=list(sample=sample))), however, I think that I am making some mistake. All my results are 0, eg.<br />Starting MCMC....<br />genearation: 1000; mean acceptance rate: 0<br />genearation: 2000; mean acceptance rate: 0<br />genearation: 3000; mean acceptance rate: 0<br /><br />And:<br />Mean correlation (r) from the posterior sample is: 0.<br />And<br />> effectiveSize(rA)<br />var1 <br /> 0 <br />> HPDinterval(rA)<br /> lower upper<br />var1 0 0<br />attr(,"Probability")<br />[1] 0.95<br /> Kaszczakhttps://www.blogger.com/profile/08932069916668207584noreply@blogger.com