NaN when creating N3LO grids for NC - F2

[scarlehoff]

I can reproduce the problem with

pinefarm run NMC_NC_EM_D_F2 4104100000.yaml

I've attached the theory card

4104100000.yaml

Looking at the grids uploaded by @ecole41 the NaNs seem to appear consistently for the F2 observable for the 00 and 03 runcars which according to the pineko docs are:

https://pineko.readthedocs.io/en/latest/theory/fonll.html

00 - ffns nf=3
03 - massive component of nf=4

Grids that are not F2 (?) but for which we get NaN:

00 - HERA_NC_318GEV_EAVG_SIGMARED_CHARM

03 - HERA_NC_318GEV_EAVG_SIGMARED_BOTTOM

[scarlehoff]

fwiw, when computing with theory 722 I get NaN as well, so it is not a problem of the definition of the theory (since we have grids at N3LO with theory 722) but something that has changed in yadism? pineappl?

[felixhekhorn]

but something that has changed in yadism? pineappl?

to explicitly check: which version did you = @scarlehoff + @ecole41 + @jekoorn use?

It is unlikely (?), but in principle also #340 could be a candidate by a wrong casting or similar (e.g. coming from an external code such as adani)

[felixhekhorn]

Actually, given these symptoms

00 - ffns nf=3
03 - massive component of nf=4

Grids that are not F2 (?) but for which we get NaN:

00 - HERA_NC_318GEV_EAVG_SIGMARED_CHARM
03 - HERA_NC_318GEV_EAVG_SIGMARED_BOTTOM

adani is a good candidate, since it is explicitly used in those

[scarlehoff]

Here's my full environment:

environment

Package                       Version                     Editable project location
----------------------------- --------------------------- -------------------------------------------------------------------
absl-py                       2.1.0
adani                         1.0.5
aiofiles                      23.2.1
alabaster                     0.7.16
anyio                         4.3.0
appdirs                       1.4.4
arcana                        0.10.16
astor                         0.8.1
asttokens                     2.4.1
astunparse                    1.6.3
attrs                         23.2.0
Babel                         2.14.0
banana-hep                    0.6.13
bcrypt                        4.1.2
bids-validator                1.14.5
black                         24.3.0
blessings                     1.7
build                         1.2.1
certifi                       2023.11.17
cffi                          1.16.0
cfgv                          3.4.0
charset-normalizer            3.3.2
ci-info                       0.3.0
click                         8.1.7
click-option-group            0.5.6
cloudpickle                   3.0.0
commonmark                    0.9.1
contourpy                     1.2.0
coverage                      7.5.0
cryptography                  42.0.2
curio                         1.6
cycler                        0.12.1
Cython                        3.0.8
dask                          2024.3.0
decorator                     5.1.1
deepdiff                      7.0.1
distlib                       0.3.8
distributed                   2024.3.0
dm-tree                       0.1.8
docker                        7.0.0
docopt                        0.6.2
docutils                      0.20.1
eko                           0.15.2
etelemetry                    0.3.1
executing                     2.0.1
fasteners                     0.19
fileformats                   0.11.2
fileformats-extras            0.11.2
filelock                      3.13.1
flatbuffers                   24.3.7
flit_core                     3.9.0
flit_scm                      1.7.0
fonttools                     4.49.0
formulaic                     0.5.2
frozendict                    2.4.0
fsspec                        2024.2.0
future                        1.0.0
gast                          0.5.4
google-pasta                  0.2.0
graphql-core                  3.2.3
greenlet                      3.0.3
grpcio                        1.62.1
h11                           0.14.0
h5py                          3.10.0
hatch-vcs                     0.4.0
hatchling                     1.24.2
httpcore                      1.0.5
httpx                         0.27.0
identify                      2.5.35
idna                          3.6
imageio                       2.34.1
imagesize                     1.4.1
importlib_metadata            7.0.2
iniconfig                     2.0.0
interface-meta                1.3.0
ipdb                          0.13.13
ipython                       8.21.0
isort                         5.13.2
jedi                          0.19.1
Jinja2                        3.1.3
joblib                        1.4.2
jsonschema                    4.22.0
jsonschema-specifications     2023.12.1
keras                         3.0.5
kiwisolver                    1.4.5
latexcodec                    3.0.0
LeProHQ                       0.2.5
lhapdf-management             0.3
libclang                      16.0.6
llvmlite                      0.44.0
locket                        1.0.0
lz4                           4.3.3
Markdown                      3.5.2
MarkupSafe                    2.1.5
matplotlib                    3.7.5
matplotlib-inline             0.1.6
medimages4tests               0.3.3
ml-dtypes                     0.3.2
more-itertools                8.14.0
msgpack                       1.0.8
mypy-extensions               1.0.0
namex                         0.0.7
networkx                      3.2.1
neurodocker                   0.9.4
nibabel                       5.2.1
nnpdf_data                    4.1.1.post35.dev0+fb8955ca8 /afs/cern.ch/work/j/jcruzmar/theory_generation/src/nnpdf/nnpdf_data
nodeenv                       1.8.0
num2words                     0.5.13
numba                         0.61.2
numpy                         2.2.6
openneuro-py                  2024.2.0
opt-einsum                    3.3.0
ordered-set                   4.1.0
packaging                     24.2
pandas                        2.3.3
paramiko                      3.4.0
parso                         0.8.3
partd                         1.4.1
pathspec                      0.12.1
pendulum                      3.0.0
pexpect                       4.9.0
pillow                        10.2.0
pineappl                      1.2.0
pinefarm                      0.3.3.post38+9ec9b05.dirty  /scratch/jcruzmar/repositories/pinefarm
pineko                        0.5.0.post10+769d0fc        /afs/cern.ch/work/j/jcruzmar/theory_generation/src/pineko
pip                           22.3.1
pkgconfig                     1.5.5
platformdirs                  4.2.0
pluggy                        1.5.0
pre-commit                    3.6.2
prompt-toolkit                3.0.43
protobuf                      4.25.3
psutil                        5.9.8
ptyprocess                    0.7.0
pure-eval                     0.2.2
py4j                          0.10.9.7
pybids                        0.16.4
pybliotecario                 2.4.1
pybtex                        0.24.0
pybtex-docutils               1.0.3
pycparser                     2.21
pydicom                       2.4.4
pydra                         0.23
pygit2                        1.14.1
Pygments                      2.18.0
PyNaCl                        1.5.0
pyparsing                     3.1.2
pyproject_hooks               1.1.0
pytest                        8.2.0
pytest-cov                    5.0.0
python-dateutil               2.8.2
pytz                          2024.1
PyYAML                        6.0.1
referencing                   0.35.1
reportengine                  0.30.dev0
requests                      2.31.0
rich                          12.6.0
rich-click                    1.9.3
rpds-py                       0.18.0
ruamel.yaml                   0.17.40
ruamel.yaml.clib              0.2.8
scipy                         1.13.0
seaborn                       0.13.2
setuptools                    65.5.1
setuptools-scm                8.0.4
sgqlc                         16.3
shellingham                   1.5.4
six                           1.16.0
sniffio                       1.3.1
snowballstemmer               2.2.0
sortedcontainers              2.4.0
SQLAlchemy                    1.4.52
stack-data                    0.6.3
tblib                         3.0.0
termcolor                     2.4.0
time-machine                  2.14.1
tomli                         2.0.1
toolz                         0.12.1
tornado                       6.4
tqdm                          4.66.2
traitlets                     5.14.1
trove-classifiers             2024.4.10
typer                         0.12.3
typing_extensions             4.10.0
tzdata                        2024.1
urllib3                       2.6.3
validobj                      1.6
virtualenv                    20.25.1
wcwidth                       0.2.13
Werkzeug                      3.0.1
wheel                         0.42.0
wrapt                         1.16.0
yadism                        0.13.9.post1+bf715b28       /scratch/jcruzmar/repositories/yadism
zict                          3.0.0
zipp                          3.18.0

adani is a good candidate, since it is explicitly used in those

meaning?

[ecole41]

but something that has changed in yadism? pineappl?

to explicitly check: which version did you = @scarlehoff + @ecole41 + @jekoorn use?

It is unlikely (?), but in principle also #340 could be a candidate by a wrong casting or similar (e.g. coming from an external code such as adani)

My environment has pineappl version 1.2.0 and yadism version 0.13.7

[felixhekhorn]

adani is a good candidate, since it is explicitly used in those

meaning?

if that would return NaN that would propagate to N3LO, but not below; However, the last tag was v1.05 on Mar 24, 2025 so it is somewhat unlikely to break now ...

pineappl?

Actually, pineappl is an unlikely candidate since that is the very last step when all numbers are already collected - and N3LO is just another step in the iteration of PTOs there ...

[scarlehoff]

if that would return NaN that would propagate to N3LO, but not below; However, the last tag was v1.05 on Mar 24, 2025 so it is somewhat unlikely to break now ...

I don't think anybody has regenerated N3LO grids since 2024 at least.

[Radonirinaunimi]

See #346 for details. This was not a problem before because $\xi$ did not reach such a high value - now it is $\xi = 10^9$.

[scarlehoff]

The preocomputed grids have NaNs that should be 0, or have NaN that by chance were set to 0?

(i.e., how much should we worry ?)