Misc changes for the NNLOJET run

.pre-commit-config.yaml

@@ -41,7 +41,7 @@ repos:
       - id: pydocstyle
         files: ^src/
         additional_dependencies:
-          - toml
+          - tomli
   - repo: https://github.com/pre-commit/pre-commit
     rev: v4.5.1
     hooks:

extras/theories/theory_nnlo.yaml

@@ -1,44 +1,44 @@
 ID: NNLO
 Comments: Example of NNLO theory
-CKM: 0.97428 0.22530 0.003470 0.22520 0.97345 0.041000 0.00862 0.04030 0.999152
+CKM: 0.97367 0.22431 3.82e-3 0.221 0.975 41.1e-3 8.6e-3 41.5e-3 1.010
 DAMP: 0
 EScaleVar: 1
 FNS: FONLL-C
-GF: 1.1663787e-05
+GF: 1.1663788e-5
 HQ: POLE
 IC: 1
 IB: 0
 MP: 0.938
-MW: 80.398
-MZ: 91.1876
+MZ: 91.1880
+MW: 80.3692
 MaxNfAs: 5
 MaxNfPdf: 5
-ModEv: TRN
+ModEv: EXA
+IterEv: 60
 ModSV: expanded
 NfFF: 4
 PTO: 2
 Q0: 1.65
 QED: 0
-Qedref: 1.777
-Qmb: 4.92
-Qmc: 1.51
-Qmt: 172.5
-Qref: 91.2
-SIN2TW: 0.23126
+Qedref: 91.1880
+Qref: 91.1880
+alphas: 0.118
+alphaqed: 0.0075700
+SIN2TW: 0.22348
 SxOrd: LL
 SxRes: 0
 TMC: 1
 XIF: 1.0
 XIR: 1.0
-alphaqed: 0.007496252
-alphas: 0.118
 global_nx: 0
-kbThr: 1.0
+mc: 1.502
+Qmc: 1.502
 kcThr: 1.0
-ktThr: 1.0
-mb: 4.92
-mc: 1.51
+mb: 4.936
+Qmb: 4.936
+kbThr: 1.0
 mt: 172.5
+Qmt: 172.5
 nfref: null
 nf0: null
 fact_to_ren_scale_ratio: 1.0

src/pinefarm/cli/autogen.py

@@ -31,10 +31,12 @@ def runcards(dataset, target, select_obs=None, name=None):
 
     output = configs.configs["paths"]["runcards"] / f"{target}_{name}"
 
-    if target == "NNLOJET":
+    if target.upper() == "NNLOJET":
         output_runcards = generate_pinecard_from_nnpdf(
             dataset, output_path=output, observables=select_obs
         )
+    else:
+        raise ValueError(f"Target {target} not recognized")
 
     rich.print("Pinecards written to: ")
     rich.print("    " + "\n".join(str(i) for i in output_runcards))

src/pinefarm/external/__init__.py

@@ -27,7 +27,7 @@ def decide_external_tool(dsname: str):
     """
     # The decisions are usually based on the existence of a `.yaml` file with a specific name
     # or a prefix in the pinecard
-    if dsname.startswith("NNLOJET"):
+    if dsname.upper().startswith("NNLOJET"):
         from .nnlojet import NNLOJET
 
         return NNLOJET, "blue"

src/pinefarm/external/nnlojet/nnpdf_interface.py

@@ -28,7 +28,7 @@
 yaml.default_flow_style = False
 yaml.indent(sequence=4, offset=2, mapping=4)
 
-HISTOGRAM_VARIABLES = {"y", "etay", "eta", "pT", "pT2", "M2"}
+HISTOGRAM_VARIABLES = {"y", "etay", "eta", "pT", "pT2", "M2", "ET"}
 
 
 def _legacy_nnpdf_translation(df, proc_type):
@@ -88,23 +88,28 @@ def _1d_histogram(kin_df, hist_var):
 def _nnlojet_observable(observable, process):
     """Try to automatically understand the NNLOJET observables given the NNPDF process and obs."""
     observable = observable.lower()
+    process = process.upper()
     if observable in ("eta", "y", "etay"):
-        if process.upper().startswith("Z"):
+        if process.startswith("Z"):
             return "yz"
-        if process.upper().startswith("WP") and not process.upper().endswith("J"):
+        if process.startswith("WP") and not process.endswith("J"):
             return "ylp"
-        if process.upper().startswith("WM") and not process.upper().endswith("J"):
+        if process.startswith("WM") and not process.endswith("J"):
             return "ylm"
+        if process.startswith("GJ"):
+            return "y_gam"
     if observable == "pt":
-        if process.upper().startswith("Z"):
+        if process.startswith("Z"):
             return "ptz"
-        if process.upper().startswith("W"):
+        if process.startswith("W"):
             return "ptw"
-    if observable == "m" and process.upper().startswith("Z"):
+    if observable == "m" and process.startswith("Z"):
         return "mll"
     if observable == "m2":
         print("\033[91m [WARNING] \033[0m Changed M2 to M in the selectors")
         return "mll"
+    if observable == "et" and process.startswith("GJ"):
+        return "pt_gam"
 
     raise ValueError(f"Observable {observable} not recognized for process {process}")
 
@@ -193,9 +198,17 @@ def _generate_nnlojet_pinecard(runname, process, energy, experiment, histograms)
     """Generate a pinecard for NNLOJET runs from an NNPDF dataset."""
     selectors = select_selectors(experiment, process)
     histograms = deepcopy(histograms)
+    photon = {}
 
+    # Digest the process variable
     if process.startswith("Z0"):
         process = process.replace("Z0", "Z")
+    if process.startswith("G"):
+        # Defaults for: 2302.00510
+        photon = {
+            "photon_isolation": "etsum[R=0.4, ETmax_const=4.8, ETmax_epsilon=0.0042, ET_threshold=0]",
+            "photon_fragmentation": "BFG2",
+        }
 
     # Digest the histogram variable
     for histo in histograms:
@@ -204,7 +217,7 @@ def _generate_nnlojet_pinecard(runname, process, energy, experiment, histograms)
 
     ret = {
         "runname": runname,
-        "process": {"proc": process, "sqrts": energy},
+        "process": {"proc": process, "sqrts": energy, **photon},
         "pdf": "NNPDF40_nnlo_as_01180",
         "techcut": 1e-7,
         "histograms": histograms,
@@ -227,11 +240,21 @@ def _generate_nnlojet_pinecard(runname, process, energy, experiment, histograms)
         },
         "selectors": selectors,
     }
+
+    # Add the scale
+    if process.startswith("Z"):
+        ret["scales"] = {"mur": "etz", "muf": "etz"}
+    elif process.startswith("W"):
+        ret["scales"] = {"mur": "etw", "muf": "etw"}
+    elif process.startswith("G"):
+        # Defaults for: 2302.00510 R=0.4
+        ret["scales"] = {"mur": "pt_gam", "muf": "[pt_gam, 0.871*sqrt_pt_gam]"}
+
     return ret
 
 
 def generate_pinecard_from_nnpdf(
-    nnpdf_dataset, scale="etz", output_path=".", observables=None
+    nnpdf_dataset, scale=None, output_path=".", observables=None
 ):
     """Generate a NNLOJET pinecard from an NNPDF dataset.
 
@@ -277,6 +300,8 @@ def generate_pinecard_from_nnpdf(
         if "M2" in hist_vars:
             if len(kin_df["M2"]["mid"].unique()) == 1:
                 hist_vars.remove("M2")
+    elif process.startswith("PH"):
+        process = process.replace("PH", "GJ")
 
     # Create the histogram depending on whether this is a 1D or 2D distribution (or total)
     histograms = None
@@ -286,10 +311,10 @@ def generate_pinecard_from_nnpdf(
     elif len(hist_vars) == 2:
 
         # Let's see whether we know how to do this 2D distribution
-        if "M2" in hist_vars:
-            svar = "M2"
-        elif "y" in hist_vars:
-            svar = "y"
+        for var_i_know in ["M2", "y", "eta"]:
+            if var_i_know in hist_vars:
+                svar = var_i_know
+                break
         else:
             raise NotImplementedError(f"Don't know how to do this 2D: {hist_vars}")
         hist_vars.remove(svar)
@@ -349,7 +374,7 @@ def generate_pinecard_from_nnpdf(
     processes = [process]
     if process.startswith("WPWM"):
         processes = [process.replace("WP", ""), process.replace("WM", "")]
-    if process == "DY":
+    elif process == "DY":
         processes = ["Z0", "WP", "WM"]
 
     parent_folder = output_path.parent
@@ -360,7 +385,9 @@ def generate_pinecard_from_nnpdf(
         runname = nnpdf_dataset.replace(process, proc)
 
         ret = _generate_nnlojet_pinecard(runname, proc, energy, experiment, histograms)
-        ret["scales"] = {"mur": scale, "muf": scale}
+        if scale is not None:
+            # Default to etz
+            ret["scales"] = {"mur": scale, "muf": scale}
 
         # Beautify before dumping
         data = CommentedMap(ret)

src/pinefarm/external/nnlojet/runcardgen.py

@@ -222,14 +222,19 @@ def parse_input_yaml(yaml_path):
 
 
 def _fill_process(process):
-    """Fill process options."""
+    """Fill process block given the metadata for the process"""
     process_name = process["proc"]
     sqrts = process["sqrts"]
-    """Fill process block given the metadata for the process"""
+    jet = process.get("jet", "none[0]")  # Can be None
+    fill_photon = ""
+    if process_name.startswith("G"):
+        fill_photon = f"""
+    photon_isolation = {process['photon_isolation']}
+    photon_fragmentation = {process['photon_fragmentation']}"""
     return f"""
 PROCESS  {process_name}
   collider = pp  sqrts = {sqrts}
-  jet = none[0]
+  jet = {jet}{fill_photon}
   decay_type = 1
 END_PROCESS
 """
@@ -239,6 +244,7 @@ def _fill_run(runname, pdf, mode_line, techcut=1e-7, multi_channel=3):
     """Fil run options."""
     if multi_channel == 0:
         multi_channel = ".false."
+    # Note, scale coefficients need to be set to true to fill the grid
     return f"""
 RUN  {runname.upper()}
   PDF = {pdf}[0]
@@ -268,6 +274,7 @@ def _fill_parameters(theory_parameters):
     return f"""
 PARAMETERS
 {ptext}
+hard_photon_alpha0 = .true. ! only useful for GJ runs
 END_PARAMETERS
 """
 

src/pinefarm/external/nnlojet/runner.py

@@ -24,6 +24,11 @@ def __init__(self, pinecard, theorycard, *args, **kwargs):
         # Save the yaml dictionary from the NNLOJET pinecard
         self._yaml_dict = safe_load(yaml_card.open("r"))
 
+    @staticmethod
+    def install():
+        """NNLOJET should be installed manually."""
+        pass
+
     def preparation(self):
         """Run the preparation step for NNLOJET."""
         # Update the yaml card according to the theory

src/pinefarm/install.py

@@ -300,7 +300,6 @@ def lhapdf():
 
     This is currently needed by every tool due to postprocessing requirements.
     """
-
     def installed():
         """Define availability condition."""
         try: