Update interpolators and add patch functionality to json2xml (#505)

- interpolator codes can be configurable for histosys similarly to normsys
- pyhf cls is updated to use the ROOT default
- remove extra (duplicate) interpolator code that got added [codex == code4p]
- pyhf json2xml now allows for patch files (#507)
- remove extraneous print statements in tests

Author: lukasheinrich

pyhf/commandline.py

@@ -3,6 +3,7 @@
 import click
 import json
 import os
+import jsonpatch
 
 from .utils import hypotest, EqDelimStringParamType
 from .pdf import Workspace
@@ -70,7 +71,8 @@ def xml2json(entrypoint_xml, basedir, output_file, track_progress):
 @click.option('--specroot', default='config')
 @click.option('--dataroot', default='data')
 @click.option('--resultprefix', default='FitConfig')
-def json2xml(workspace, output_dir, specroot, dataroot, resultprefix):
+@click.option('-p', '--patch', multiple=True)
+def json2xml(workspace, output_dir, specroot, dataroot, resultprefix, patch):
     try:
         import uproot
 
@@ -86,6 +88,9 @@ def json2xml(workspace, output_dir, specroot, dataroot, resultprefix):
     ensure_dirs(output_dir)
     with click.open_file(workspace, 'r') as specstream:
         d = json.load(specstream)
+        for pfile in patch:
+            patch = json.loads(click.open_file(pfile, 'r').read())
+            d = jsonpatch.JsonPatch(patch).apply(d)
         ensure_dirs(os.path.join(output_dir, specroot))
         ensure_dirs(os.path.join(output_dir, dataroot))
         with click.open_file(
@@ -238,7 +243,10 @@ def cls(
     p = w.model(
         measurement_name=measurement,
         patches=patches,
-        modifier_settings={'normsys': {'interpcode': 'code4'}},
+        modifier_settings={
+            'normsys': {'interpcode': 'code4'},
+            'histosys': {'interpcode': 'code4p'},
+        },
     )
 
     optconf = {k: v for item in optconf for k, v in item.items()}

pyhf/interpolators/__init__.py

@@ -39,4 +39,4 @@ def get(interpcode, do_tensorized_calc=True):
         raise exceptions.InvalidInterpCode
 
 
-__all__ = ['code0', 'code1', 'code2']
+__all__ = ['code0', 'code1', 'code2', 'code4', 'code4p']

pyhf/interpolators/codex.py

@@ -25,8 +25,10 @@ def required_parset(cls, n_parameters):
 
 
 class histosys_combined(object):
-    def __init__(self, histosys_mods, pdfconfig, mega_mods):
+    def __init__(self, histosys_mods, pdfconfig, mega_mods, interpcode='code0'):
         self._parindices = list(range(len(pdfconfig.suggested_init())))
+        self.interpcode = interpcode
+        assert self.interpcode in ['code0', 'code2', 'code4p']
 
         pnames = [pname for pname, _ in histosys_mods]
         keys = ['{}/{}'.format(mtype, m) for m, mtype in histosys_mods]
@@ -48,7 +50,9 @@ def __init__(self, histosys_mods, pdfconfig, mega_mods):
         ]
 
         if len(histosys_mods):
-            self.interpolator = interpolators.code0(self._histosys_histoset)
+            self.interpolator = getattr(interpolators, self.interpcode)(
+                self._histosys_histoset
+            )
 
         self._precompute()
         events.subscribe('tensorlib_changed')(self._precompute)

pyhf/modifiers/histosys.py

@@ -333,9 +333,9 @@ def _create_nominal_and_modifiers(self):
         self.thenom = default_backend.reshape(
             thenom,
             (
-                1,
+                1,  # modifier dimension.. thenom is the base
                 len(self.config.samples),
-                1,
+                1,  # alphaset dimension
                 sum(list(self.config.channel_nbins.values())),
             ),
         )

pyhf/pdf.py

@@ -130,6 +130,13 @@ def test_patch(tmpdir, script_runner):
     command = 'pyhf cls {0:s} --patch {1:s}'.format(temp.strpath, patch.strpath)
     ret = script_runner.run(*shlex.split(command))
     assert ret.success
+
+    command = 'pyhf json2xml {0:s} --output-dir {1:s} --patch {2:s}'.format(
+        temp.strpath, tmpdir.mkdir('output_1').strpath, patch.strpath
+    )
+    ret = script_runner.run(*shlex.split(command))
+    assert ret.success
+
     import io
 
     command = 'pyhf cls {0:s} --patch -'.format(temp.strpath, patch.strpath)
@@ -138,7 +145,15 @@ def test_patch(tmpdir, script_runner):
         patchcontent
     )  # python 2.7 pytest-files are not file-like enough
     ret = script_runner.run(*shlex.split(command), stdin=pipefile)
-    print(ret.stderr)
+    assert ret.success
+
+    command = 'pyhf json2xml {0:s} --output-dir {1:s} --patch -'.format(
+        temp.strpath, tmpdir.mkdir('output_2').strpath, patch.strpath
+    )
+    pipefile = io.StringIO(
+        patchcontent
+    )  # python 2.7 pytest-files are not file-like enough
+    ret = script_runner.run(*shlex.split(command), stdin=pipefile)
     assert ret.success
 
 
@@ -157,6 +172,12 @@ def test_patch_fail(tmpdir, script_runner):
     ret = script_runner.run(*shlex.split(command))
     assert not ret.success
 
+    command = 'pyhf json2xml {0:s} --output-dir {1:s} --patch {2:s}'.format(
+        temp.strpath, tmpdir.mkdir('output').strpath, patch.strpath
+    )
+    ret = script_runner.run(*shlex.split(command))
+    assert not ret.success
+
 
 def test_bad_measurement_name(tmpdir, script_runner):
     temp = tmpdir.join("parsed_output.json")