Page 1 of 1

Trajectory display (seemingly) not producing a file.

Posted: December 2nd, 2020, 4:07 pm
by cm922
Hello everyone,

I've set up TCL/TK and got hysplit working, and installed the relevant Anaconda3 packages that should allow the Trajectory Display program to output with Python to pdf. However, when I click "execute display", the following comes up:

C:\hysplit\working>"C:/Users/[NAME]/anaconda3/envs/hysplit/../../Scripts/activate.bat" "C:/Users/[NAME]/anaconda3/envs/hysplit" && python.exe "C:/hysplit/exec/trajplot.py" --interactive --debug -iC:/hysplit/working/TrajectoryOutput/RE2008_02122020_ -otrajplot.pdf -jC:/hysplit/graphics/arlmap -m1 -k1 -l6 -z50 -v1 -a3 -s1 -g4:100 -h-74:-78.0

and it seems like no pdf file of the trajectory is created, certainly not one I can find at least. Has anyone else had this issue/is there an easy fix?

Thank you all,

-C

Re: Trajectory display (seemingly) not producing a file.

Posted: December 3rd, 2020, 8:26 am
by sonny.zinn
The problem might be caused by the pandas python package. The Hysplit Python graphics works well with pandas 1.0.1 but it does not with 1.1.1. Please see viewtopic.php?p=6069#p6069.

Re: Trajectory display (seemingly) not producing a file.

Posted: December 4th, 2020, 1:30 pm
by cm922
Thank you for the reply.

I did what you mentioned in that thread and it throws up this error:

ImportError: cannot import name '_NDFrameIndexer' from 'pandas.core.indexing' (C:\Users\Cian\anaconda3\envs\hysplit\lib\site-packages\pandas-1.1.4-py3.7-win-amd64.egg\pandas\core\indexing.py)

Do I need to change to pandas 1.0.1, and if so how would I do that?

Re: Trajectory display (seemingly) not producing a file.

Posted: December 4th, 2020, 2:09 pm
by sonny.zinn
The following worked with my Windows 10 laptop. Open an Anaconda Prompt. Then run:

> conda activate hysplit
> python -m pip install pandas==1.0.1

Let me know if this resolved the Python graphics issue.

Re: Trajectory display (seemingly) not producing a file.

Posted: December 4th, 2020, 7:25 pm
by cm922
This works and produces a figure. Thank you so much for your help!

Re: Trajectory display (seemingly) not producing a file.

Posted: December 4th, 2020, 10:00 pm
by cm922
Apologies for a follow up problem, but I'm now trying to use the Trajectory Frequency Plot program.

Daily run program works and produces the files in C:\hysplit\working
INFILE can be created from these in the Trajectory Frequency Plot program.
Program can be executed, and it creates the initial freqplot.pdf in C:\hysplit\working

However, the simulation never finishes, ending up stuck with the following simulation log indefinitely:
Gridding: trajfreq -ftfreq.bin -g1.0 -iINFILE -r0 -s500:3000 -b000000000001
Input files processed: 120
Count sum over entire domain: 2499.000
Total number of traj: 120
Total endpoint count: 5760
Endpoints/Trajectory: 48
Complete Trajfreq
and also a second log with thousands of mainly font related statements:
01:47:19.121 DEBUG matplotlib.font_manager - findfont: score(<Font 'Kristen ITC' (ITCKRIST.TTF) normal normal 400 normal>) = 10.05
01:47:19.127 DEBUG matplotlib.font_manager - findfont: score(<Font 'Franklin Gothic Heavy' (FRAHVIT.TTF) italic normal heavy normal>) = 11.43
01:47:19.127 DEBUG matplotlib.font_manager - findfont: score(<Font 'Microsoft YaHei' (msyhl.ttc) normal normal light normal>) = 10.24
01:47:19.128 DEBUG matplotlib.font_manager - findfont: score(<Font 'Calisto MT' (CALISTBI.TTF) italic normal bold normal>) = 11.335
...
01:47:22.528 DEBUG matplotlib.font_manager - findfont: Matching :family=DejaVu Sans Display:style=normal:variant=normal:weight=normal:stretch=normal:size=10.0 to DejaVu Sans Display ('C:\\Users\\[NAME]\\anaconda3\\envs\\hysplit\\lib\\site-packages\\matplotlib\\mpl-data\\fonts\\ttf\\DejaVuSansDisplay.ttf') with score of 0.050000.
The produced pdf file can't be opened (I assume it's not been written to by the program as it only has a size of 1KB). This particular run was done with trajectories every six hours for a month, but even just a file with trajectories every hour for a day caused the same issue.