Segmentation fault in clusterEvents

[bluemood2011]

Hi,

Thanks for providing such a great tool for alternative splicing analysis.
Recently I was doing clusterEvents for both differential events and transcripts, but encountered a 'Segmentation fault' issue for transcripts for several times, which is weird because my data is not very large. I checked the output and can found all the clusters in dtu.clustvec.

Here is my command:
suppa.py clusterEvents -st 0.05 -dt 0.05 -e 0.05 -m euclidean -n 2 -c DBSCAN -d dtu.dpsi -p dtu.psivec -g 1-3,4-6 -o dtu
Number of clustered events: 1358/1368 (99.269006%)
Cluster 0 contains 642 events
Cluster 1 contains 657 events
Cluster 2 contains 2 events
Cluster 3 contains 8 events
Cluster 4 contains 5 events
Cluster 5 contains 2 events
Cluster 6 contains 9 events
Cluster 7 contains 2 events
Cluster 8 contains 4 events
Cluster 9 contains 3 events
Cluster 10 contains 3 events
Cluster 11 contains 3 events
Cluster 12 contains 7 events
Cluster 13 contains 3 events
Cluster 14 contains 3 events
Cluster 15 contains 5 events
The RMSSTD score for cluster 0 is 0.707023
The RMSSTD score for cluster 1 is 0.710955
The RMSSTD score for cluster 2 is 0.176591
The RMSSTD score for cluster 3 is 0.000000
The RMSSTD score for cluster 4 is 0.000000
The RMSSTD score for cluster 5 is 0.000000
The RMSSTD score for cluster 6 is 0.274745
The RMSSTD score for cluster 7 is 0.165720
The RMSSTD score for cluster 8 is 0.000000
The RMSSTD score for cluster 9 is 0.125951
The RMSSTD score for cluster 10 is 0.142581
The RMSSTD score for cluster 11 is 0.142581
The RMSSTD score for cluster 12 is 0.235498
The RMSSTD score for cluster 13 is 0.133740
The RMSSTD score for cluster 14 is 0.085284
The RMSSTD score for cluster 15 is 0.080613
Segmentation fault