Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Caching of Non-Equilibrium parameters(Evib, Erot) #511

Open
wants to merge 16 commits into
base: develop
Choose a base branch
from
Open
Changes from 2 commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
80 changes: 65 additions & 15 deletions radis/test/lbl/test_base.py
Original file line number Diff line number Diff line change
Expand Up @@ -5,6 +5,7 @@
@author: erwan
"""

import os
import time
from os.path import exists, getmtime, splitext

Expand Down Expand Up @@ -831,56 +832,105 @@ def test_caching_noneq_params(verbose=True, plot=True, *args, **kwargs):
verbose=2,
)
s.fetch_databank("hitemp", load_energies=True, load_columns="noneq")
s1 = s.non_eq_spectrum(Tvib=2000, Trot=3000)
s2 = s.non_eq_spectrum(Tvib=2000, Trot=3000)
# Loading variables
molecule = s.input.molecule
isotope = s.input.isotope
state = "X"
# Checking that the cache file exists
# Checking that the cache file exists, removing if it does
cache_filename = (
splitext(s.params.dbpath.split(",", 1)[0])[0]
+ "_extra_columns_EvibErot.h5.extra"
)
engine = "pytables"
if exists(cache_filename):
last_modif = getmtime(cache_filename)
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

shouldn't last_modif be measured after computing the new spectrum, i.e.> l846 ?

os.remove(cache_filename)
# Generating first spectrum
s1 = s.non_eq_spectrum(Tvib=2000, Trot=3000)
assert exists(cache_filename)
# Generating second spectrum using another SpectrumFactory object:
sf = SpectrumFactory(
wavelength_min=1500,
wavelength_max=2000,
cutoff=1e-27,
pressure=1,
molecule="CO",
isotope="1,2",
truncation=5,
neighbour_lines=5,
path_length=0.1,
mole_fraction=1e-3,
medium="vacuum",
optimization=None, # No optimization strategy used
chunksize=None, # Initialising chunksize as None for comparison
wstep="auto",
verbose=2,
Comment on lines +849 to +863
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

to simplify reading, can you add all of this in a dictionary and use the same dictionary many times.

conditions = {
        "wavelength_min":1500,
        "wavelength_max":2000,
#etc, fix for dictionary syntax : 
        cutoff=1e-27,
        pressure=1,
        molecule="CO",
        isotope="1,2",
        truncation=5,
        neighbour_lines=5,
        path_length=0.1,
        mole_fraction=1e-3,
        medium="vacuum",
        optimization=None,  # No optimization strategy used
        chunksize=None,  # Initialising chunksize as None for comparison
        wstep="auto",
        verbose=2,
}

s1 = SpectrumFactory(**conditions)
s2 = SpectrumFactory(**conditions)
# or change one : 
new_conditions = conditions.copy()
new_conditions["wavelength_min=1800"]
s3 = SpectrumFactory(**new_conditions)

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Also, remove all parametesr that are not relevant. pressure, Path_length, mole_fraction, medium, optimization, chunksize, wstep, verbose are not relevant here. Keep the defaults

)
sf.fetch_databank("hitemp", load_energies=True, load_columns="noneq")
s2 = sf.non_eq_spectrum(Tvib=2000, Trot=3000)
# Checking proper regeneration of the cache file
assert exists(cache_filename)
assert getmtime(cache_filename) > last_modif
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

It seems to me the conditions are the same. Shouldn't the modification time be the same (i.e. file is not regenerated)


cache_file_data = DataFileManager(engine).read(cache_filename)
# Checking that the dataframe non-equilibrium columns are the same
assert sf.df0["Evibu"].equals(cache_file_data["Evibu"])
assert sf.df0["Evibl"].equals(cache_file_data["Evibl"])
assert sf.df0["Erotu"].equals(cache_file_data["Erotu"])
assert sf.df0["Erotl"].equals(cache_file_data["Erotl"])

# Loading variables
molecule = sf.input.molecule
isotope = sf.input.isotope
state = "X"

elec_state = s.get_partition_function_calculator(
molecule, s._get_isotope_list(molecule)[0], state
).ElecState
spectroscopic_constant_file = elec_state.jsonfile

# Checking that the metadata is correct, else regenerating the cache file
existing_file_metadata = DataFileManager(engine).read_metadata(cache_filename)

# Checking that the metadata is correct
if "isotope" in existing_file_metadata:
assert isotope == existing_file_metadata["isotope"]
else:
os.remove(cache_filename)
if "number_of_lines" in existing_file_metadata:
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Wait. Why are you testing this here ??

These validity checks should be in the code itself, not in the test.

Here you should only be testing (and asserting) that if you change isotope, the cache file is properly regenerated (ie new_modif_time > last_modif_time)

Same for wavenumbers, etc.

assert len(s.df0) == existing_file_metadata["number_of_lines"]
assert len(sf.df0) == existing_file_metadata["number_of_lines"]
else:
os.remove(cache_filename)
if "wavenum_min" in existing_file_metadata:
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Test should happen always. If key is not in metadata you should regenerate the cache

assert s.input.wavenum_min == existing_file_metadata["wavenum_min"]
assert sf.input.wavenum_min == existing_file_metadata["wavenum_min"]
else:
os.remove(cache_filename)
if "wavenum_max" in existing_file_metadata:
assert s.input.wavenum_max == existing_file_metadata["wavenum_max"]
assert sf.input.wavenum_max == existing_file_metadata["wavenum_max"]
else:
os.remove(cache_filename)
if "spectroscopic_constant_file" in existing_file_metadata:
assert (
spectroscopic_constant_file
== existing_file_metadata["spectroscopic_constant_file"]
)
else:
os.remove(cache_filename)
if "last_modification" in existing_file_metadata:
assert (
time.ctime(getmtime(spectroscopic_constant_file))
== existing_file_metadata["last_modification"]
)
else:
os.remove(cache_filename)
if "neighbour_lines" in existing_file_metadata:
assert s.params.neighbour_lines == existing_file_metadata["neighbour_lines"]
assert sf.params.neighbour_lines == existing_file_metadata["neighbour_lines"]
else:
os.remove(cache_filename)
if "cutoff" in existing_file_metadata:
assert s.params.cutoff == existing_file_metadata["cutoff"]
assert sf.params.cutoff == existing_file_metadata["cutoff"]
else:
os.remove(cache_filename)
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

same as comment above. You shouldn't be conditionning the presence of "last_modifcation" etc in the metadata.

It must be there. You should even be asserting that it is there.


res = get_residual(s1, s2, "radiance_noslit")
if verbose:
print(
"Res for non-equilibrium spectrum with molecule = {0}, wavenumber range = {1} to {2}: {3}".format(
s.input.molecule, s.df0["wav"].min(), s.df0["wav"].max(), res
s.input.molecule, s.input.wavenum_min, s.input.wavenum_max, res
)
)
assert res < 1e-6
Expand Down