File size: 32,542 Bytes
5980447 |
1 2 |
{"repo": "gammapy/gammapy", "pull_number": 2296, "instance_id": "gammapy__gammapy-2296", "issue_numbers": "", "base_commit": "a7dbdc9598e531b4089a707375b434a567795e3c", "patch": "diff --git a/gammapy/cube/models.py b/gammapy/cube/models.py\n--- a/gammapy/cube/models.py\n+++ b/gammapy/cube/models.py\n@@ -1,11 +1,12 @@\n # Licensed under a 3-clause BSD style license - see LICENSE.rst\n import copy\n+from pathlib import Path\n import numpy as np\n import astropy.units as u\n from ..utils.fitting import Parameter, Model, Parameters\n-from ..utils.scripts import make_path\n from ..spectrum.models import SpectralModel\n from ..image.models import SkySpatialModel\n+from ..utils.scripts import make_path, write_yaml\n from ..maps import Map\n \n __all__ = [\n@@ -120,6 +121,21 @@ def to_xml(self, filename):\n with filename.open(\"w\") as output:\n output.write(xml)\n \n+ @classmethod\n+ def from_yaml(cls, filename):\n+ \"\"\"Write to YAML file.\"\"\"\n+ from ..utils.serialization import dict_to_models\n+ from ..utils.scripts import read_yaml\n+ data = read_yaml(filename)\n+ skymodels = dict_to_models(data)\n+ return cls(skymodels)\n+\n+ def to_yaml(self, filename, selection=\"all\"):\n+ \"\"\"Write to YAML file.\"\"\"\n+ from ..utils.serialization import models_to_dict\n+ components_dict = models_to_dict(self.skymodels, selection)\n+ write_yaml(components_dict, filename)\n+\n def evaluate(self, lon, lat, energy):\n out = self.skymodels[0].evaluate(lon, lat, energy)\n for skymodel in self.skymodels[1:]:\n@@ -299,7 +315,16 @@ class SkyDiffuseCube(SkyModelBase):\n \n __slots__ = [\"map\", \"norm\", \"meta\", \"_interp_kwargs\"]\n \n- def __init__(self, map, norm=1, meta=None, interp_kwargs=None, name=\"diffuse\"):\n+ def __init__(\n+ self,\n+ map,\n+ norm=1,\n+ meta=None,\n+ interp_kwargs=None,\n+ name=\"diffuse\",\n+ filename=None,\n+ obs_id=\"Global\",\n+ ):\n self.name = name\n axis = map.geom.get_axis_by_name(\"energy\")\n \n@@ -309,6 +334,8 @@ def __init__(self, map, norm=1, meta=None, interp_kwargs=None, name=\"diffuse\"):\n self.map = map\n self.norm = Parameter(\"norm\", norm)\n self.meta = {} if meta is None else meta\n+ self.filename = filename\n+ self.obs_id = obs_id\n \n interp_kwargs = {} if interp_kwargs is None else interp_kwargs\n interp_kwargs.setdefault(\"interp\", \"linear\")\n@@ -331,7 +358,8 @@ def read(cls, filename, **kwargs):\n m = Map.read(filename, **kwargs)\n if m.unit == \"\":\n m.unit = \"cm-2 s-1 MeV-1 sr-1\"\n- return cls(m)\n+ name = Path(filename).stem\n+ return cls(m, name=name, filename=filename)\n \n def evaluate(self, lon, lat, energy):\n \"\"\"Evaluate model.\"\"\"\n@@ -381,9 +409,18 @@ class BackgroundModel(Model):\n Reference energy of the tilt.\n \"\"\"\n \n- __slots__ = [\"map\", \"norm\", \"tilt\", \"reference\"]\n-\n- def __init__(self, background, norm=1, tilt=0, reference=\"1 TeV\"):\n+ __slots__ = [\"map\", \"norm\", \"tilt\", \"reference\", \"name\", \"filename\", \"obs_id\"]\n+\n+ def __init__(\n+ self,\n+ background,\n+ norm=1,\n+ tilt=0,\n+ reference=\"1 TeV\",\n+ name=\"background\",\n+ filename=None,\n+ obs_id=None,\n+ ):\n axis = background.geom.get_axis_by_name(\"energy\")\n if axis.node_type != \"edges\":\n raise ValueError('Need an integrated map, energy axis node_type=\"edges\"')\n@@ -392,7 +429,9 @@ def __init__(self, background, norm=1, tilt=0, reference=\"1 TeV\"):\n self.norm = Parameter(\"norm\", norm, unit=\"\", min=0)\n self.tilt = Parameter(\"tilt\", tilt, unit=\"\", frozen=True)\n self.reference = Parameter(\"reference\", reference, frozen=True)\n-\n+ self.name = name\n+ self.filename = filename\n+ self.obs_id = obs_id\n super().__init__([self.norm, self.tilt, self.reference])\n \n @property\n@@ -440,7 +479,12 @@ def from_skymodel(cls, skymodel, exposure, edisp=None, psf=None, **kwargs):\n model=skymodel, exposure=exposure, edisp=edisp, psf=psf\n )\n background = evaluator.compute_npred()\n- return cls(background=background, **kwargs)\n+ background_model = cls(background=background, **kwargs)\n+ background_model.name = skymodel.name\n+ background_model.obs_id = skymodel.obs_id\n+ if skymodel.__class__.__name__ == \"SkyDiffuseCube\":\n+ background_model.filename = skymodel.filename\n+ return background_model\n \n def __add__(self, model):\n models = [self]\n@@ -494,3 +538,9 @@ def __add__(self, model):\n model_ = self.copy()\n model_ += model\n return model_\n+\n+ def to_yaml(self, filename, selection=\"all\"):\n+ \"\"\"Write to yaml file.\"\"\"\n+ from ..utils.serialization import models_to_dict\n+ components_dict = models_to_dict(self.models, selection)\n+ write_yaml(components_dict, filename)\ndiff --git a/gammapy/data/obs_stats.py b/gammapy/data/obs_stats.py\n--- a/gammapy/data/obs_stats.py\n+++ b/gammapy/data/obs_stats.py\n@@ -139,7 +139,7 @@ def stack(cls, stats_list):\n def to_dict(self):\n \"\"\"Data as a dict.\n \n- This is useful for serialisation or putting the info in a table.\n+ This is useful for serialization or putting the info in a table.\n \"\"\"\n return {\n \"obs_id\": self.obs_id,\ndiff --git a/gammapy/image/models/core.py b/gammapy/image/models/core.py\n--- a/gammapy/image/models/core.py\n+++ b/gammapy/image/models/core.py\n@@ -534,9 +534,11 @@ class SkyDiffuseMap(SkySpatialModel):\n Default arguments are {'interp': 'linear', 'fill_value': 0}.\n \"\"\"\n \n- __slots__ = [\"map\", \"norm\", \"meta\", \"_interp_kwargs\"]\n+ __slots__ = [\"map\", \"norm\", \"meta\", \"_interp_kwargs\", \"filename\"]\n \n- def __init__(self, map, norm=1, meta=None, normalize=True, interp_kwargs=None):\n+ def __init__(\n+ self, map, norm=1, meta=None, normalize=True, interp_kwargs=None, filename=None\n+ ):\n if (map.data < 0).any():\n log.warning(\"Diffuse map has negative values. Check and fix this!\")\n \n@@ -552,7 +554,7 @@ def __init__(self, map, norm=1, meta=None, normalize=True, interp_kwargs=None):\n interp_kwargs.setdefault(\"interp\", \"linear\")\n interp_kwargs.setdefault(\"fill_value\", 0)\n self._interp_kwargs = interp_kwargs\n-\n+ self.filename = filename\n super().__init__([self.norm])\n \n @property\n@@ -587,7 +589,7 @@ def read(cls, filename, normalize=True, **kwargs):\n m = Map.read(filename, **kwargs)\n if m.unit == \"\":\n m.unit = \"sr-1\"\n- return cls(m, normalize=normalize)\n+ return cls(m, normalize=normalize, filename=filename)\n \n def evaluate(self, lon, lat, norm):\n \"\"\"Evaluate model.\"\"\"\ndiff --git a/gammapy/spectrum/models.py b/gammapy/spectrum/models.py\n--- a/gammapy/spectrum/models.py\n+++ b/gammapy/spectrum/models.py\n@@ -212,18 +212,12 @@ def f(x):\n uarray = integrate_spectrum(f, emin.value, emax.value, **kwargs)\n return self._parse_uarray(uarray) * unit\n \n- def to_dict(self):\n- \"\"\"Convert to dict.\"\"\"\n- retval = self.parameters.to_dict()\n- retval[\"name\"] = self.__class__.__name__\n- return retval\n-\n @classmethod\n- def from_dict(cls, val):\n+ def from_dict(cls, data):\n \"\"\"Create from dict.\"\"\"\n- val_copy = val.copy()\n- classname = val_copy.pop(\"name\")\n- parameters = Parameters.from_dict(val_copy)\n+ data = data.copy()\n+ classname = data.pop(\"type\")\n+ parameters = Parameters.from_dict(data)\n model = globals()[classname]()\n model.parameters = parameters\n model.parameters.covariance = parameters.covariance\n@@ -1327,6 +1321,20 @@ def evaluate(self, energy, norm):\n values = self._evaluate((energy,), clip=True)\n return norm * values\n \n+ def to_dict(self, selection=\"all\"):\n+ return {\n+ \"type\": self.__class__.__name__,\n+ \"parameters\": self.parameters.to_dict(selection)[\"parameters\"],\n+ \"energy\": {\n+ \"data\": self.energy.data.tolist(),\n+ \"unit\": str(self.energy.unit),\n+ },\n+ \"values\": {\n+ \"data\": self.values.data.tolist(),\n+ \"unit\": str(self.values.unit),\n+ },\n+ }\n+\n \n class ScaleModel(SpectralModel):\n \"\"\"Wrapper to scale another spectral model by a norm factor.\ndiff --git a/gammapy/utils/fits.py b/gammapy/utils/fits.py\n--- a/gammapy/utils/fits.py\n+++ b/gammapy/utils/fits.py\n@@ -130,7 +130,7 @@\n \n >>> hdu = fits.table_to_hdu(table)\n \n-However, in this case, the column metadata that is serialised is\n+However, in this case, the column metadata that is serialized is\n doesn't include the column ``description``.\n TODO: how to get consistent behaviour and FITS headers?\n \ndiff --git a/gammapy/utils/fitting/model.py b/gammapy/utils/fitting/model.py\n--- a/gammapy/utils/fitting/model.py\n+++ b/gammapy/utils/fitting/model.py\n@@ -38,3 +38,9 @@ def __str__(self):\n covariance = self.parameters.covariance_to_table()\n ss += \"\\n\\t\".join(covariance.pformat())\n return ss\n+\n+ def to_dict(self, selection=\"all\"):\n+ return {\n+ \"type\": self.__class__.__name__,\n+ \"parameters\": self.parameters.to_dict(selection)[\"parameters\"],\n+ }\ndiff --git a/gammapy/utils/fitting/parameter.py b/gammapy/utils/fitting/parameter.py\n--- a/gammapy/utils/fitting/parameter.py\n+++ b/gammapy/utils/fitting/parameter.py\n@@ -179,17 +179,34 @@ def __repr__(self):\n \"min={min!r}, max={max!r}, frozen={frozen!r})\"\n ).format(**self.to_dict())\n \n- def to_dict(self):\n- return dict(\n- name=self.name,\n- value=self.value,\n- factor=self.factor,\n- scale=self.scale,\n- unit=self.unit.to_string(\"fits\"),\n- min=self.min,\n- max=self.max,\n- frozen=self.frozen,\n- )\n+ def to_dict(self, selection=\"all\"):\n+ \"\"\"Convert to dict.\n+\n+ Parameters\n+ -----------\n+ selection : {\"all\", \"simple\"}\n+ Selection of information to include\n+ \"\"\"\n+ if selection == \"simple\":\n+ return dict(\n+ name=self.name,\n+ value=self.value,\n+ unit=self.unit.to_string(\"fits\"),\n+ frozen=self.frozen,\n+ )\n+ elif selection == \"all\":\n+ return dict(\n+ name=self.name,\n+ value=self.value,\n+ factor=self.factor,\n+ scale=self.scale,\n+ unit=self.unit.to_string(\"fits\"),\n+ min=self.min,\n+ max=self.max,\n+ frozen=self.frozen,\n+ )\n+ else:\n+ raise ValueError(\"Invalid selection: {!r}\".format(selection))\n \n def autoscale(self, method=\"scale10\"):\n \"\"\"Autoscale the parameters.\n@@ -330,13 +347,14 @@ def __getitem__(self, name):\n idx = self._get_idx(name)\n return self.parameters[idx]\n \n- def to_dict(self):\n- retval = dict(parameters=[], covariance=None)\n+ def to_dict(self, selection=\"all\"):\n+ data = dict(parameters=[], covariance=None)\n for par in self.parameters:\n- retval[\"parameters\"].append(par.to_dict())\n+ data[\"parameters\"].append(par.to_dict(selection))\n if self.covariance is not None:\n- retval[\"covariance\"] = self.covariance.tolist()\n- return retval\n+ data[\"covariance\"] = self.covariance.tolist()\n+\n+ return data\n \n def to_table(self):\n \"\"\"Convert parameter attributes to `~astropy.table.Table`.\"\"\"\n@@ -359,25 +377,25 @@ def to_table(self):\n return t\n \n @classmethod\n- def from_dict(cls, val):\n- pars = []\n- for par in val[\"parameters\"]:\n- pars.append(\n- Parameter(\n- name=par[\"name\"],\n- factor=float(par[\"value\"]),\n- unit=par[\"unit\"],\n- min=float(par[\"min\"]),\n- max=float(par[\"max\"]),\n- frozen=par[\"frozen\"],\n- )\n+ def from_dict(cls, data):\n+ parameters = []\n+ for par in data[\"parameters\"]:\n+ parameter = Parameter(\n+ name=par[\"name\"],\n+ factor=float(par[\"value\"]),\n+ unit=par.get(\"unit\", \"\"),\n+ min=float(par.get(\"min\", np.nan)),\n+ max=float(par.get(\"max\", np.nan)),\n+ frozen=par.get(\"frozen\", False),\n )\n+ parameters.append(parameter)\n+\n try:\n- covariance = np.array(val[\"covariance\"])\n+ covariance = np.array(data[\"covariance\"])\n except KeyError:\n covariance = None\n \n- return cls(parameters=pars, covariance=covariance)\n+ return cls(parameters=parameters, covariance=covariance)\n \n def covariance_to_table(self):\n \"\"\"Convert covariance matrix to `~astropy.table.Table`.\"\"\"\ndiff --git a/gammapy/utils/scripts.py b/gammapy/utils/scripts.py\n--- a/gammapy/utils/scripts.py\n+++ b/gammapy/utils/scripts.py\n@@ -43,21 +43,26 @@ def _configure_root_logger(level=\"info\", format=None):\n \n \n def read_yaml(filename, logger=None):\n- \"\"\"\n- Read YAML file\n+ \"\"\"Read YAML file.\n \n Parameters\n ----------\n- filename : `pathlib.Path`, str\n- File to read\n+ filename : `~pathlib.Path`\n+ Filename\n+ logger : `~logging.Logger`\n+ Logger\n+\n+ Returns\n+ -------\n+ data : dict\n+ YAML file content as a dict\n \"\"\"\n- filename = make_path(filename)\n+ path = make_path(filename)\n if logger is not None:\n- logger.info(\"Reading {}\".format(filename))\n- with open(str(filename)) as fh:\n- dictionary = yaml.safe_load(fh)\n+ logger.info(\"Reading {}\".format(path))\n \n- return dictionary\n+ text = path.read_text()\n+ return yaml.safe_load(text)\n \n \n def write_yaml(dictionary, filename, logger=None):\n@@ -67,15 +72,18 @@ def write_yaml(dictionary, filename, logger=None):\n ----------\n dictionary : dict\n Python dictionary\n- filename : str, `~gammapy.exter.pathlib.Path`\n- file to write\n+ filename : `~pathlib.Path`\n+ Filename\n+ logger : `~logging.Logger`\n+ Logger\n \"\"\"\n- filename = make_path(filename)\n- filename.parent.mkdir(exist_ok=True)\n+ text = yaml.safe_dump(dictionary, default_flow_style=False)\n+\n+ path = make_path(filename)\n+ path.parent.mkdir(exist_ok=True)\n if logger is not None:\n- logger.info(\"Writing {}\".format(filename))\n- with open(str(filename), \"w\") as outfile:\n- outfile.write(yaml.safe_dump(dictionary, default_flow_style=False))\n+ logger.info(\"Writing {}\".format(path))\n+ path.write_text(text)\n \n \n def make_path(path):\ndiff --git a/gammapy/utils/serialization/__init__.py b/gammapy/utils/serialization/__init__.py\n--- a/gammapy/utils/serialization/__init__.py\n+++ b/gammapy/utils/serialization/__init__.py\n@@ -1,3 +1,4 @@\n \"\"\"Serialization utility functions.\n \"\"\"\n from .xml import *\n+from .io import *\ndiff --git a/gammapy/utils/serialization/io.py b/gammapy/utils/serialization/io.py\nnew file mode 100644\n--- /dev/null\n+++ b/gammapy/utils/serialization/io.py\n@@ -0,0 +1,111 @@\n+# Licensed under a 3-clause BSD style license - see LICENSE.rst\n+\"\"\"Utilities to serialize models.\"\"\"\n+import astropy.units as u\n+from ...image import models as spatial\n+from ...spectrum import models as spectral\n+from ...cube.models import SkyModel\n+from ..fitting import Parameters\n+\n+__all__ = [\n+ \"models_to_dict\",\n+ \"dict_to_models\",\n+]\n+\n+\n+def models_to_dict(models, selection=\"all\"):\n+ \"\"\"Convert list of models to dict.\n+\n+ Parameters\n+ -----------\n+ models : list\n+ Python list of Model objects\n+ selection : {\"all\", \"simple\"}\n+ Selection of information to include\n+ \"\"\"\n+ models_data = []\n+ for model in models:\n+ model_data = _model_to_dict(model, selection)\n+\n+ # De-duplicate if model appears several times\n+ if model_data not in models_data:\n+ models_data.append(model_data)\n+\n+ return {\"components\": models_data}\n+\n+\n+def _model_to_dict(model, selection):\n+ data = {}\n+ data[\"name\"] = getattr(model, \"name\", model.__class__.__name__)\n+ try:\n+ data[\"id\"] = model.obs_id\n+ except AttributeError:\n+ pass\n+ if getattr(model, \"filename\", None) is not None:\n+ data[\"filename\"] = model.filename\n+ if model.__class__.__name__ == \"SkyModel\":\n+ data[\"spatial\"] = model.spatial_model.to_dict(selection)\n+ if getattr(model.spatial_model, \"filename\", None) is not None:\n+ data[\"spatial\"][\"filename\"] = model.spatial_model.filename\n+ data[\"spectral\"] = model.spectral_model.to_dict(selection)\n+ else:\n+ data[\"model\"] = model.to_dict(selection)\n+\n+ return data\n+\n+\n+def dict_to_models(data):\n+ \"\"\"De-serialise model data to Model objects.\n+\n+ Parameters\n+ -----------\n+ data : dict\n+ Serialised model information\n+ \"\"\"\n+ models = []\n+ for model in data[\"components\"]:\n+ if \"model\" in model:\n+ if model[\"model\"][\"type\"] == \"BackgroundModel\":\n+ continue\n+ else:\n+ raise NotImplementedError\n+\n+ model = _dict_to_skymodel(model)\n+ models.append(model)\n+\n+ return models\n+\n+\n+def _dict_to_skymodel(model):\n+ item = model[\"spatial\"]\n+ if \"filename\" in item:\n+ spatial_model = getattr(spatial, item[\"type\"]).read(item[\"filename\"])\n+ spatial_model.filename = item[\"filename\"]\n+ spatial_model.parameters = Parameters.from_dict(item)\n+ else:\n+ params = {\n+ x[\"name\"]: x[\"value\"] * u.Unit(x[\"unit\"])\n+ for x in item[\"parameters\"]\n+ }\n+ spatial_model = getattr(spatial, item[\"type\"])(**params)\n+ spatial_model.parameters = Parameters.from_dict(item)\n+\n+ item = model[\"spectral\"]\n+ if \"energy\" in item:\n+ energy = u.Quantity(item[\"energy\"][\"data\"], item[\"energy\"][\"unit\"])\n+ values = u.Quantity(item[\"values\"][\"data\"], item[\"values\"][\"unit\"])\n+ params = {\"energy\": energy, \"values\": values}\n+ spectral_model = getattr(spectral, item[\"type\"])(**params)\n+ spectral_model.parameters = Parameters.from_dict(item)\n+ else:\n+ params = {\n+ x[\"name\"]: x[\"value\"] * u.Unit(x[\"unit\"])\n+ for x in item[\"parameters\"]\n+ }\n+ spectral_model = getattr(spectral, item[\"type\"])(**params)\n+ spectral_model.parameters = Parameters.from_dict(item)\n+\n+ return SkyModel(\n+ name=model[\"name\"],\n+ spatial_model=spatial_model,\n+ spectral_model=spectral_model,\n+ )\ndiff --git a/gammapy/utils/setup_package.py b/gammapy/utils/setup_package.py\nnew file mode 100644\n--- /dev/null\n+++ b/gammapy/utils/setup_package.py\n@@ -0,0 +1,6 @@\n+# Licensed under a 3-clause BSD style license - see LICENSE.rst\n+\n+\n+def get_package_data():\n+ files = [\"data/*\"]\n+ return {\"gammapy.utils.serialization.tests\": files}\n", "test_patch": "diff --git a/gammapy/catalog/tests/test_snrcat.py b/gammapy/catalog/tests/test_snrcat.py\n--- a/gammapy/catalog/tests/test_snrcat.py\n+++ b/gammapy/catalog/tests/test_snrcat.py\n@@ -13,7 +13,7 @@ def test_load_catalog_snrcat(tmpdir):\n assert len(table) > 300\n expected_colnames = [\"Source_Name\", \"RAJ2000\"]\n assert set(expected_colnames).issubset(table.colnames)\n- # Check if catalog can be serialised to FITS\n+ # Check if catalog can be serialized to FITS\n filename = str(tmpdir / \"snrcat_test.fits\")\n table.write(filename)\n \n@@ -23,6 +23,6 @@ def test_load_catalog_snrcat(tmpdir):\n expected_colnames = [\"SNR_id\", \"source_id\"]\n assert set(expected_colnames).issubset(table.colnames)\n \n- # Check if catalog can be serialised to FITS\n+ # Check if catalog can be serialized to FITS\n filename = str(tmpdir / \"obs_test.fits\")\n table.write(filename)\ndiff --git a/gammapy/maps/tests/test_wcs.py b/gammapy/maps/tests/test_wcs.py\n--- a/gammapy/maps/tests/test_wcs.py\n+++ b/gammapy/maps/tests/test_wcs.py\n@@ -393,7 +393,7 @@ def test_wcs_geom_equal(npix, binsz, coordsys, proj, skypos, axes, result):\n @pytest.mark.parametrize(\"node_type\", [\"edges\", \"center\"])\n @pytest.mark.parametrize(\"interp\", [\"log\", \"lin\", \"sqrt\"])\n def test_read_write(tmpdir, node_type, interp):\n- # Regression test for MapAxis interp and node_type FITS serialisation\n+ # Regression test for MapAxis interp and node_type FITS serialization\n # https://github.com/gammapy/gammapy/issues/1887\n e_ax = MapAxis([1, 2], interp, \"energy\", node_type, \"TeV\")\n t_ax = MapAxis([3, 4], interp, \"time\", node_type, \"s\")\ndiff --git a/gammapy/utils/serialization/tests/data/example2.yaml b/gammapy/utils/serialization/tests/data/example2.yaml\nnew file mode 100644\n--- /dev/null\n+++ b/gammapy/utils/serialization/tests/data/example2.yaml\n@@ -0,0 +1,56 @@\n+components:\n+- name: source\n+ spatial:\n+ parameters:\n+ - factor: 0.0\n+ frozen: false\n+ max: 180.0\n+ min: -180.0\n+ name: lon_0\n+ scale: 1.0\n+ unit: deg\n+ value: 0.0\n+ - factor: 0.0\n+ frozen: false\n+ max: 90.0\n+ min: -90.0\n+ name: lat_0\n+ scale: 1.0\n+ unit: deg\n+ value: 0.0\n+ - factor: 1.0\n+ frozen: false\n+ max: .nan\n+ min: 0.0\n+ name: sigma\n+ scale: 1.0\n+ unit: deg\n+ value: 1.0\n+ type: SkyGaussian\n+ spectral:\n+ parameters:\n+ - factor: 2.0\n+ frozen: false\n+ max: .nan\n+ min: .nan\n+ name: index\n+ scale: 1.0\n+ unit: ''\n+ value: 2.0\n+ - factor: 1.0e-12\n+ frozen: false\n+ max: .nan\n+ min: .nan\n+ name: amplitude\n+ scale: 1.0\n+ unit: cm-2 s-1 TeV-1\n+ value: 1.0e-12\n+ - factor: 1.0\n+ frozen: true\n+ max: .nan\n+ min: .nan\n+ name: reference\n+ scale: 1.0\n+ unit: TeV\n+ value: 1.0\n+ type: PowerLaw\ndiff --git a/gammapy/utils/serialization/tests/data/examples.yaml b/gammapy/utils/serialization/tests/data/examples.yaml\nnew file mode 100644\n--- /dev/null\n+++ b/gammapy/utils/serialization/tests/data/examples.yaml\n@@ -0,0 +1,190 @@\n+components:\n+- name: source0\n+ spatial:\n+ type: SkyPointSource\n+ parameters:\n+ - name: lon_0\n+ value: -50.\n+ factor: -500.\n+ scale: 0.01\n+ unit: deg\n+ min: -180.0\n+ max: 180.0\n+ frozen: true\n+ - name: lat_0\n+ value: -0.05\n+ factor: -5.\n+ scale: 0.01\n+ unit: deg\n+ min: -90.0\n+ max: 90.0\n+ frozen: true\n+ spectral:\n+ type: ExponentialCutoffPowerLaw\n+ parameters:\n+ - name: index\n+ value: 2.1\n+ factor: 2.1\n+ scale: 1.0\n+ unit: ''\n+ min: .nan\n+ max: .nan\n+ frozen: false\n+ - name: amplitude\n+ value: 2.3e-12\n+ factor: 2.3\n+ scale: 1.0e-12\n+ unit: cm-2 s-1 TeV-1\n+ min: .nan\n+ max: .nan\n+ frozen: false\n+ - name: reference\n+ value: 1.0\n+ factor: 1.0\n+ scale: 1.0\n+ unit: TeV\n+ min: .nan\n+ max: .nan\n+ frozen: true\n+ - name: lambda_\n+ value: 0.06\n+ factor: 0.6\n+ scale: 0.1\n+ unit: TeV-1\n+ min: .nan\n+ max: .nan\n+ frozen: false\n+- name: source1\n+ spatial:\n+ type: SkyDisk\n+ parameters:\n+ - name: lon_0\n+ value: -50.\n+ unit: deg\n+ frozen: false\n+ - name: lat_0\n+ value: -0.05\n+ unit: deg\n+ frozen: false\n+ - name: r_0\n+ value: 0.2\n+ unit: deg\n+ frozen: false\n+ spectral:\n+ type: PowerLaw\n+ parameters:\n+ - name: index\n+ value: 2.2\n+ factor: 2.2\n+ scale: 1.0\n+ unit: ''\n+ min: .nan\n+ max: .nan\n+ frozen: false\n+ - name: amplitude\n+ value: 2.3e-12\n+ factor: 2.3\n+ scale: 1.0e-12\n+ unit: cm-2 s-1 TeV-1\n+ min: .nan\n+ max: .nan\n+ frozen: false\n+ - name: reference\n+ value: 1.0\n+ factor: 1.0\n+ scale: 1.0\n+ unit: TeV\n+ min: .nan\n+ max: .nan\n+ frozen: true\n+- name: source2\n+ spatial:\n+ type: SkyDiffuseMap\n+ filename: $GAMMAPY_DATA/catalogs/fermi/Extended_archive_v18/Templates/RXJ1713_2016_250GeV.fits\n+ parameters:\n+ - name: norm\n+ value: 1.0\n+ unit: ''\n+ frozen: false\n+ spectral:\n+ type: TableModel\n+ energy:\n+ data:\n+ - 34.171\n+ - 44.333\n+ - 57.517\n+ unit: MeV\n+ values:\n+ data:\n+ - 2.52894e-06\n+ - 1.2486e-06\n+ - 6.14648e-06\n+ unit: 1 / (cm2 MeV s sr)\n+ parameters:\n+ - name: norm\n+ value: 2.1\n+ factor: 2.1\n+ scale: 1.0\n+ unit: ''\n+ min: .nan\n+ max: .nan\n+ frozen: false\n+- name: background_irf\n+ id: CTA-gc\n+ model:\n+ type: BackgroundModel\n+ parameters:\n+ - name: norm\n+ value: 1.01\n+ factor: 1.01\n+ scale: 1.0\n+ unit: ''\n+ min: 0.0\n+ max: .nan\n+ frozen: false\n+ - name: tilt\n+ value: 0.0\n+ factor: 0.0\n+ scale: 1.0\n+ unit: ''\n+ min: .nan\n+ max: .nan\n+ frozen: true\n+ - name: reference\n+ value: 1.0\n+ factor: 1.0\n+ scale: 1.0\n+ unit: TeV\n+ min: .nan\n+ max: .nan\n+ frozen: true\n+- name: cube_iem\n+ id: global\n+ filename: $GAMMAPY_DATA/fermi-3fhl/gll_iem_v06_cutout.fits\n+ model:\n+ type: BackgroundModel\n+ parameters:\n+ - name: norm\n+ value: 1.09\n+ factor: 1.09\n+ scale: 1.0\n+ unit: ''\n+ min: 0.0\n+ max: .nan\n+ frozen: false\n+ - name: tilt\n+ value: 0.0\n+ factor: 0.0\n+ scale: 1.0\n+ unit: ''\n+ min: .nan\n+ max: .nan\n+ frozen: true\n+ - name: reference\n+ value: 1.0\n+ factor: 1.0\n+ scale: 1.0\n+ unit: TeV\n+ min: .nan\n+ max: .nan\n+ frozen: true\ndiff --git a/gammapy/utils/serialization/tests/make_test_data.py b/gammapy/utils/serialization/tests/make_test_data.py\nnew file mode 100644\n--- /dev/null\n+++ b/gammapy/utils/serialization/tests/make_test_data.py\n@@ -0,0 +1,21 @@\n+\"\"\"Create example model YAML files programmatically.\n+\n+(some will be also written manually)\n+\"\"\"\n+from pathlib import Path\n+from gammapy.image.models import SkyGaussian\n+from gammapy.spectrum.models import PowerLaw\n+from gammapy.cube.models import SkyModels, SkyModel\n+\n+DATA_PATH = Path(\"gammapy/utils/serialization/tests/data/\")\n+\n+\n+def make_example_2():\n+ spatial = SkyGaussian(\"0 deg\", \"0 deg\", \"1 deg\")\n+ model = SkyModel(spatial, PowerLaw())\n+ models = SkyModels([model])\n+ models.to_yaml(DATA_PATH / \"example2.yaml\")\n+\n+\n+if __name__ == '__main__':\n+ make_example_2()\ndiff --git a/gammapy/utils/serialization/tests/test_io.py b/gammapy/utils/serialization/tests/test_io.py\nnew file mode 100644\n--- /dev/null\n+++ b/gammapy/utils/serialization/tests/test_io.py\n@@ -0,0 +1,100 @@\n+# Licensed under a 3-clause BSD style license - see LICENSE.rst\n+import numpy as np\n+from numpy.testing import assert_allclose\n+from astropy.utils.data import get_pkg_data_filename\n+from ...testing import requires_data\n+from ....spectrum import models as spectral\n+from ....image import models as spatial\n+from ....cube.models import SkyModels\n+from ...scripts import read_yaml, write_yaml\n+from ...serialization import models_to_dict, dict_to_models\n+\n+\n+@requires_data()\n+def test_dict_to_skymodels(tmpdir):\n+ filename = get_pkg_data_filename(\"data/examples.yaml\")\n+ models_data = read_yaml(filename)\n+ models = dict_to_models(models_data)\n+\n+ assert len(models) == 3\n+\n+ model0 = models[0]\n+ assert isinstance(model0.spectral_model, spectral.ExponentialCutoffPowerLaw)\n+ assert isinstance(model0.spatial_model, spatial.SkyPointSource)\n+\n+ pars0 = model0.parameters\n+ assert pars0[\"index\"].value == 2.1\n+ assert pars0[\"index\"].unit == \"\"\n+ assert np.isnan(pars0[\"index\"].max)\n+ assert np.isnan(pars0[\"index\"].min)\n+ assert pars0[\"index\"].frozen is False\n+\n+ assert pars0[\"lon_0\"].value == -50.0\n+ assert pars0[\"lon_0\"].unit == \"deg\"\n+ assert pars0[\"lon_0\"].max == 180.0\n+ assert pars0[\"lon_0\"].min == -180.0\n+ assert pars0[\"lon_0\"].frozen is True\n+\n+ assert pars0[\"lat_0\"].value == -0.05\n+ assert pars0[\"lat_0\"].unit == \"deg\"\n+ assert pars0[\"lat_0\"].max == 90.0\n+ assert pars0[\"lat_0\"].min == -90.0\n+ assert pars0[\"lat_0\"].frozen is True\n+\n+ assert pars0[\"lambda_\"].value == 0.06\n+ assert pars0[\"lambda_\"].unit == \"TeV-1\"\n+ assert np.isnan(pars0[\"lambda_\"].min)\n+ assert np.isnan(pars0[\"lambda_\"].max)\n+\n+ model1 = models[1]\n+ assert isinstance(model1.spectral_model, spectral.PowerLaw)\n+ assert isinstance(model1.spatial_model, spatial.SkyDisk)\n+\n+ pars1 = model1.parameters\n+ assert pars1[\"index\"].value == 2.2\n+ assert pars1[\"index\"].unit == \"\"\n+ assert pars1[\"lat_0\"].scale == 1.0\n+ assert pars1[\"lat_0\"].factor == pars1[\"lat_0\"].value\n+\n+ assert np.isnan(pars1[\"index\"].max)\n+ assert np.isnan(pars1[\"index\"].min)\n+\n+ assert pars1[\"r_0\"].unit == \"deg\"\n+\n+ model2 = models[2]\n+ assert_allclose(model2.spectral_model.energy.data, [34.171, 44.333, 57.517])\n+ assert model2.spectral_model.energy.unit == \"MeV\"\n+ assert_allclose(\n+ model2.spectral_model.values.data, [2.52894e-06, 1.2486e-06, 6.14648e-06]\n+ )\n+ assert model2.spectral_model.values.unit == \"1 / (cm2 MeV s sr)\"\n+\n+ assert isinstance(model2.spectral_model, spectral.TableModel)\n+ assert isinstance(model2.spatial_model, spatial.SkyDiffuseMap)\n+\n+ assert model2.spatial_model.parameters[\"norm\"].value == 1.0\n+ assert model2.spectral_model.parameters[\"norm\"].value == 2.1\n+ # TODO problem of duplicate parameter name between SkyDiffuseMap and TableModel\n+ # assert model2.parameters[\"norm\"].value == 2.1 # fail\n+\n+\n+# TODO: test background model serialisation\n+\n+@requires_data()\n+def test_sky_models_io(tmpdir):\n+ # TODO: maybe change to a test case where we create a model programatically?\n+ filename = get_pkg_data_filename(\"data/examples.yaml\")\n+ models = SkyModels.from_yaml(filename)\n+\n+ filename = str(tmpdir / \"io_example.yaml\")\n+ models.to_yaml(filename)\n+ SkyModels.from_yaml(filename)\n+ # TODO: add asserts to check content\n+\n+ models.to_yaml(filename, selection=\"simple\")\n+ SkyModels.from_yaml(filename)\n+ # TODO add assert to check content\n+\n+ # TODO: not sure if we should just round-trip, or if we should\n+ # check YAML file content (e.g. against a ref file in the repo)\n+ # or check serialised dict content\n", "problem_statement": "", "hints_text": "", "created_at": "2019-07-18T12:35:17Z"}
|