def test_baseobj_get_parameters(setup_pars: dict):
name = setup_pars["name"]
del setup_pars["name"]
obj = BaseObj(name, **setup_pars)
pars = obj.get_fit_parameters()
assert isinstance(pars, list)
assert len(pars) == 2
par_names = [par.name for par in pars]
assert "p2" in par_names
assert "p3" in par_names
def test_baseobj__add_component(setup_pars):
name = setup_pars["name"]
del setup_pars["name"]
obj = BaseObj(name, **setup_pars)
p = Parameter("added_par", 1)
new_item_name = "Added"
obj._add_component(new_item_name, p)
assert hasattr(obj, new_item_name)
a = getattr(obj, new_item_name)
assert isinstance(a, Parameter)
def test_baseobj_set(setup_pars: dict):
from copy import deepcopy
name = setup_pars["name"]
explicit_name1 = "par1"
kwargs = {
setup_pars[explicit_name1].name: setup_pars[explicit_name1],
}
obj = BaseObj(name, **kwargs)
new_value = 5.0
with not_raises([AttributeError, ValueError]):
obj.p1 = new_value
assert obj.p1.raw_value == new_value
def test_baseobj_dir(setup_pars):
name = setup_pars["name"]
del setup_pars["name"]
obj = BaseObj(name, **setup_pars)
expected = [
"REDIRECT",
"as_dict",
"constraints",
"des1",
"des2",
"from_dict",
"generate_bindings",
"get_fit_parameters",
"get_parameters",
"interface",
"name",
"par1",
"par2",
"par3",
"switch_interface",
"to_data_dict",
"to_json",
"unsafe_hash",
"user_data",
]
obtained = dir(obj)
assert len(obtained) == len(expected)
assert obtained == sorted(obtained)
for this_item, this_expect in zip(obtained, expected):
assert this_item == this_expect
def test_baseCollection_get_fit_parameters_nested(cls, setup_pars):
name = setup_pars['name']
del setup_pars['name']
obj = BaseObj(name, **setup_pars)
name2 = name + '_2'
obj2 = cls(name2, obj=obj, **setup_pars)
pars = obj2.get_fit_parameters()
assert len(pars) == 4
def test_baseobj_get(setup_pars: dict):
name = setup_pars["name"]
explicit_name1 = "par1"
explicit_name2 = "par2"
kwargs = {
setup_pars[explicit_name1].name: setup_pars[explicit_name1],
setup_pars[explicit_name2].name: setup_pars[explicit_name2],
}
obj = BaseObj(name, **kwargs)
with not_raises(AttributeError):
p1: Parameter = obj.p1
with not_raises(AttributeError):
p2: Parameter = obj.p2
def test_baseobj_create(setup_pars: dict, a: List[str], kw: List[str]):
name = setup_pars["name"]
args = []
for key in a:
args.append(setup_pars[key])
kwargs = {}
for key in kw:
kwargs[key] = setup_pars[key]
base = BaseObj(name, *args, **kwargs)
assert base.name == name
for key in a:
item = getattr(base, setup_pars[key].name)
assert isinstance(item, setup_pars[key].__class__)
def test_BaseObjUndoRedo():
objs = {
obj.name: obj
for obj in [createSingleObjs(idx) for idx in range(5)]
}
name = 'test'
obj = BaseObj(name, **objs)
name2 = 'best'
# Test name
# assert not doUndoRedo(obj, 'name', name2)
# Test setting value
for b_obj in objs.values():
e = doUndoRedo(obj, b_obj.name, b_obj.raw_value + 1, 'raw_value')
if e:
raise e
def test_baseobj_get_fit_parameters_nested(setup_pars):
name = setup_pars["name"]
del setup_pars["name"]
obj = BaseObj(name, **setup_pars)
name2 = name + "_2"
obj2 = BaseObj(name2, obj=obj, **setup_pars)
pars = obj2.get_fit_parameters()
assert len(pars) == 4
pars = obj.get_fit_parameters()
assert len(pars) == 2
def test_baseCollection_from_baseObj(cls, setup_pars: dict, value: int):
name = setup_pars['name']
del setup_pars['name']
objs = {}
prefix = 'obj'
for idx in range(value):
objs[prefix + str(idx)] = BaseObj(prefix + str(idx), **setup_pars)
coll = cls(name, **objs)
assert coll.name == name
assert len(coll) == value
assert coll.user_data == {}
idx = 0
for item, key in zip(coll, objs.keys()):
assert item.name == prefix + str(idx)
assert isinstance(item, objs[key].__class__)
idx += 1
__author__ = 'github.com/wardsimon'
__version__ = '0.1.0'
from easyCore import np
from easyCore.Datasets.xarray import xr
from easyCore.Objects.ObjectClasses import Parameter, BaseObj
from easyCore.Fitting.Fitting import Fitter
import matplotlib.pyplot as plt
d = xr.Dataset()
b = BaseObj('line', m=Parameter('m', 1), c=Parameter('c', 1))
def fit_fun(x, *args, **kwargs):
# In the real case we would gust call the evaluation fn without reference to the BaseObj
return b.c.raw_value + b.m.raw_value * x
f = Fitter()
f.initialize(b, fit_fun)
nx = 1E3
x_min = 0
x_max = 100
x = np.linspace(x_min, x_max, num=int(nx))
y1 = 2 * x - 1 + 5 * (np.random.random(size=x.shape) - 0.5)
x2 = x + 20
y2 = 2 * x2 - 1 + 5 * (np.random.random(size=x2.shape) - 0.5)
def test_baseObj_name(setup_pars):
name = setup_pars["name"]
del setup_pars["name"]
obj = BaseObj(name, **setup_pars)
assert obj.name == name
def test_baseobj_get_fit_parameters(setup_pars):
name = setup_pars["name"]
del setup_pars["name"]
obj = BaseObj(name, **setup_pars)
pars = obj.get_fit_parameters()
assert len(pars) == 2
def test_baseobj_as_dict(setup_pars: dict):
name = setup_pars["name"]
del setup_pars["name"]
obj = BaseObj(name, **setup_pars)
obtained = obj.as_dict()
assert isinstance(obtained, dict)
expected = {
"@module": "easyCore.Objects.ObjectClasses",
"@class": "BaseObj",
"@version": easyCore.__version__,
"name": "test",
"par1": {
"@module": "easyCore.Objects.Variable",
"@class": "Parameter",
"@version": easyCore.__version__,
"name": "p1",
"value": 0.1,
"error": 0.0,
"min": -np.inf,
"max": np.inf,
"fixed": True,
"units": "dimensionless",
},
"des1": {
"@module": "easyCore.Objects.Variable",
"@class": "Descriptor",
"@version": easyCore.__version__,
"name": "d1",
"value": 0.1,
"units": "dimensionless",
"description": "",
"url": "",
"display_name": "d1",
},
"par2": {
"@module": "easyCore.Objects.Variable",
"@class": "Parameter",
"@version": easyCore.__version__,
"name": "p2",
"value": 0.1,
"error": 0.0,
"min": -np.inf,
"max": np.inf,
"fixed": False,
"units": "dimensionless",
},
"des2": {
"@module": "easyCore.Objects.Variable",
"@class": "Descriptor",
"@version": easyCore.__version__,
"name": "d2",
"value": 0.1,
"units": "dimensionless",
"description": "",
"url": "",
"display_name": "d2",
},
"par3": {
"@module": "easyCore.Objects.Variable",
"@class": "Parameter",
"@version": easyCore.__version__,
"name": "p3",
"value": 0.1,
"error": 0.0,
"min": -np.inf,
"max": np.inf,
"fixed": False,
"units": "dimensionless",
},
}
def check_dict(check, item):
if isinstance(check, dict) and isinstance(item, dict):
if "@module" in item.keys():
with not_raises([ValueError, AttributeError]):
this_obj = MontyDecoder().process_decoded(item)
for this_check_key, this_item_key in zip(check.keys(), item.keys()):
check_dict(check[this_check_key], item[this_item_key])
else:
assert isinstance(item, type(check))
assert item == check
check_dict(expected, obtained)
__version__ = '0.1.0'
from easyCore import np
from easyCore.Datasets.xarray import xr
import matplotlib.pyplot as plt
from easyCore.Objects.ObjectClasses import Parameter, BaseObj
from easyCore.Fitting.Fitting import Fitter
d = xr.Dataset()
m_starting_point = 1
c_starting_point = 1
b = BaseObj('line',
m=Parameter('m', m_starting_point),
c=Parameter('c', c_starting_point))
def fit_fun(x, *args, **kwargs):
# In the real case we would gust call the evaluation fn without reference to the BaseObj
return b.c.raw_value + b.m.raw_value * x
nx = 1E3
x_min = 0
x_max = 100
x = np.linspace(x_min, x_max, num=int(nx))
y = 2*x - 1 + 5*(np.random.random(size=x.shape) - 0.5)
x = np.linspace(x_min, x_max, num=int(nx))
d.easyCore.add_coordinate('x', x)
d.easyCore.add_coordinate('y', x + 1)
d.easyCore.add_variable(
'z', ['x', 'y'],
np.sin(2 * np.pi * (x + sin_offest)).reshape(
(-1, 1)) * np.cos(2 * np.pi * (x + cos_offset)).reshape(
(1, -1)) + 2 * (0.5 - np.random.random(size=(int(nx), int(nx)))))
s_off_start_point = 0
c_off_start_point = 0
b = BaseObj('line',
s_off=Parameter('s_off', s_off_start_point),
c_off=Parameter('c_off', c_off_start_point))
def fit_fun(x, *args, **kwargs):
# In the real case we would gust call the evaluation fn without reference to the BaseObj
return np.sin(2 * np.pi * (x[:, 0] + b.s_off.raw_value)) * np.cos(
2 * np.pi * (x[:, 1] + b.c_off.raw_value))
f = Fitter()
f.initialize(b, fit_fun)
print('Performing fit - No dask')
t = time.time()
f_res = d['z'].easyCore.fit(f)