Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
| From | "Stefan Weih" <Stefan.Weih@wiso.stud.uni-erlangen.de> |
| To | statalist@hsphsun2.harvard.edu |
| Subject | Re: st: Clustering with Sequence analysis/optimal matching |
| Date | Wed, 21 Mar 2012 21:49:45 +0100 |
Thanks for your comments, Brendan. However, as indicated in the correspondance earlier, I did use Ward's method. Also no typo involved. For the complete syntax on my clustering procedure, please see below: sqclusterdat clustermat wardslinkage SQdist, name(wards) add cluster tree wards, cutnumber(20) sqclusterdat, return Thanks in advance for any further feedback. Best regards Stefan _____________________ > According to my limited understanding of cluster analysis, reversals > shouldn't occur with Ward's method. They are expected with some other > methods. > Did you actually use Ward's method? > Is your distance matrix actually called wards or is that a typo in your -cluster tree- command? > Brendan -- > Brendan Halpin, Department of Sociology, University of Limerick, > Ireland > Tel: w +353-61-213147 f +353-61-202569 h +353-61-338562; Room F1-009 x > 3147 > mailto:brendan.halpin@ul.ie ULSociology on Facebook: http://on.fb.me/fjIK9t http://teaching.sociology.ul.ie/bhalpin/wordpress twitter:@ULSociology ________________________ Uli, thank you very much. The update works well regarding the clustering. Please, let me ask one more question: In determining the optimal number of clusters I tried to create a dendrogram using the syntax: cluster tree wards, cutnumber(20) Unfortunately, it returns the error message "currently can't handle dendrogram reversals". Am I missing something here? Or is there are a way to make it work? Alternatively, is there another command which I could apply? I am thinking of e.g. cluster stop (however, I am not quite sure which variables then to use with this command). Highly appreciate your support. Best regards Stefan ____________________________ > Stefan, > > the problem that you are reporting typcially happens with the version of > the program that was distributed togehter with the cited SJ article. > There is a newer version of the program available on SSC. Please use > > . ado uninstall st0111 > . ssc install sq > > to update to the most recent version of the package. > > Note as an aside that the most recent version of the sq-package also has > a lot of new features, including further functions for -egen-, a way to > apply multidimensional scaling to the distance matrix, decreased > overplotting of lines in sequence index plots, and a new graphical > device for plotting "ideal typical" sequences (-sqmodalplot-). > > Many regards > > Uli > > > > Am Mittwoch, den 21.03.2012, 11:38 +0100 schrieb Stefan Weih: >> Dear Statalist-members, >> >> I am currently analyzing diversification strategy sequences with 16 >> potential sequence states using the sqom (optimal matching) command. >> As a common next step I would like to cluster my sqom-results. For this, >> I >> am using the command sqclusterdat procedure as described in ?Sequence >> analysis with Stata? (Brzinsky-Fay/Kohler/Luniak, 2006, The Stata >> Journal6(4), 435-460). The syntax I am using looks the following: >> >> sqset Seq ID Order >> >> sqom, subcost(matrixA) full >> >> sqclusterdat >> clustermat wardslinkage SQdist, name(wards) add >> sqclusterdat, return >> >> However, the clustering does not work properly. After performing >> sqclusterdat, return I keep receiving the following message ?Group >> results >> could not be merged to sequence data. Returned to original sequence >> data?. >> >> Unfortunately the cause for this message is not obvious to me. I am >> thankful for any guidance and hints that can help me out. >> >> Sincerely, >> Stefan Weih >> >> >> The data looks like: >> >> ID Seq Order >> 41 11 1 >> 41 13 2 >> 41 11 3 >> 41 21 4 >> 42 11 1 >> 42 13 2 >> 42 12 3 >> 42 11 4 >> 43 14 1 >> 43 13 2 >> 43 11 3 >> 43 12 4 >> >> >> * >> * For searches and help try: >> * http://www.stata.com/help.cgi?search >> * http://www.stata.com/support/statalist/faq >> * http://www.ats.ucla.edu/stat/stata/ > > > > * > * For searches and help try: > * http://www.stata.com/help.cgi?search > * http://www.stata.com/support/statalist/faq > * http://www.ats.ucla.edu/stat/stata/ > * * For searches and help try: * http://www.stata.com/help.cgi?search * http://www.stata.com/support/statalist/faq * http://www.ats.ucla.edu/stat/stata/