nuspacesim/nupyprop

nupyprop's Issues

0 km water depth gives wrong Pexit for small angles-testing mode

Multi-energy runs of nupyprop launch infinite-energy simulations

Running 2 energies as command line arguments causes the code to fail to parse the energy list and instead run infinite energy, which never completes.

$ nupyprop -a 1,3 -e 1e7,1e8                                                                                                                        
The water -> rock transition occurs at 2.03 degrees
Energy = inf, Angle = 1.0

When fixed, be sure to include a test case in the tests directory to catch this type of failure moving forward.

file eout for -htc yes fails for angle .ge. 10

When -htc yes is set, two files should be written. For -e 10 (10^10 GeV energy) -a 1 (1 degree), the files are
pexit_10.00_1.00.dat
eout_10.00_1.00.dat
The bug is when -a 10. The files for the same -e 10 are
pexit_10.00_10.00.dat
eout_10.00_****.dat

pexit output bug for series of runs with same statistics with default -htc

for this series of runs of nupyprop
nupyprop -e 8 -a 1
nupyprop -e 8 -a 2
nupyprop -e 8 -a 3,4
The output file output_nu_tau_4km_ct18nlo_allm_stochastic_1e7.h5 has:
CLep_out_cdf has a group for -e 8 (8.0) with columns of CDFs for all 4 angles
CLep_out_energies has a group 8.0 which has 4 groups (1.0,2.0,3.0,4.0), one for each angle
Exit_Probability has one group 8.0 with only 2 rows, one for 3. and one for 4.

With statistics, models etc the same, except for angles, the cdf and energies are appended with each subsequent call to angles, but the pexit group is overwritten, so the results for -a 1 and -a 2 are lost.

Anomalous errors when running on HTC cluster

README file fails twine validation checks

Not sure what the real issue here is, but the existing README.md fails to parse correctly during the twine validation checks.
Twine is necessary to upload wheels to pypi.

The actual error is:

$ python -m twine check nupyprop-0.1.3-cp39-cp39-linux_x86_64.whl                                                                                                 
Checking nupyprop-0.1.3-cp39-cp39-linux_x86_64.whl: FAILED
  `long_description` has syntax errors in markup and would not be rendered on PyPI.
    line 70: Error: Undefined substitution referenced: "---".
  warning: `long_description_content_type` missing. defaulting to `text/x-rst`.

Default arguments are not working

Running without any command line arguments causes nupyprop to perform no computation. Either the defaults should be computed if no arguments are given, or the help output and other documentation should be updated to reflect the new lack of defaults.

$ nupyprop 
Done in 2.03 seconds!

When fixed, be sure to include a test case in the tests directory to catch this type of failure moving forward.

Deprecate python 3.7 support

Remove support for python 3.7

program failure for -a 9.99

command line: nupyprop -e 8 -a 9.99 -t stochastic -s 1e+6 -htc yes
errors:
File "/Users/reno/opt/anaconda3/envs/nupyprop/bin/nupyprop", line 33, in
sys.exit(load_entry_point('nupyprop', 'console_scripts', 'nupyprop')())
File "/Users/reno/Work/NuSpaceSim/nupyprop/src/nupyprop/scripts/run.py", line 174, in main
Main.main(
File "/Users/reno/Work/NuSpaceSim/nupyprop/src/nupyprop/main.py", line 193, in main
chord, water = Data.get_trajs('water', angle, idepth)
File "/Users/reno/Work/NuSpaceSim/nupyprop/src/nupyprop/data.py", line 221, in get_trajs
chord = float(traj_table['chord'][traj_table['beta']==angle])
TypeError: only size-1 arrays can be converted to Python scalars

importlib_resources is creating problems when trying to load the lookup_tables.h5

Some issues with different python versions and importlib_resources. This may be expected given the lack of support for importlib.resources before 3.9

Correcting fortran code for all rock, i.e., 0 km