Error while generating Mean cluster in Trajectory Cluster analysis

Topics about the trajectory clustering program for HYSPLIT.
Post Reply
indumati_sp
Posts: 22
Joined: November 20th, 2012, 12:46 am
Registered HYSPLIT User: Yes

Error while generating Mean cluster in Trajectory Cluster analysis

Post by indumati_sp »

I get the following error message while creating mean trajectory through "Display Means" though the "Display clusters" generates individual trajectory clusters.

"NOTICE: If input files have different number of diagnostic variables, and the file with the most is not first in the input list, the program will fail without any other error message."

Kindly advice.
TIA
Harris_MRP
Posts: 1
Joined: September 3rd, 2022, 12:43 am
Registered HYSPLIT User: No

Re: Error while generating Mean cluster in Trajectory Cluster analysis

Post by Harris_MRP »

Hey mate, just posting as I was getting an identical error and I managed to solve it. I had been trying to plot the cluster means for a set of endpoints that contained different combinations of output variables (i.e. pressure, theta, air temp, rainfall, etc.). The clusters were not running, so I wrote some code that removed all output variables from the endpoint files except for pressure. After this, the clusters would successfully calculate, but the mean plotting failed.

It turns out the cause of this is a mismatch between the record #5 I6 element (see https://www.ready.noaa.gov/hysplitusersguide/S263.htm), which (I think) tells the clustering program how many output variables a given endpoint file contains, and the actual number of output variables contained in the endpoint file.

I revised my code so that the record #5 I6 element matched the number of diagnostic variables in each modified endpoint file. At this point the means plotted successfully.

So, this mean cluster plotting error is likely caused by either (1) a difference in the number of diagnostic output variables within the endpoints files you used for the cluster calculation, or (2) a mismatch between the record #5 I6 element and the actual number of diagnostic variables in one or more of your endpoint files.

Hope this helps.
Post Reply

Return to “Cluster Analysis”