From a259c9ac88ef2ac88a2e8ef6b358e5bf221d21e8 Mon Sep 17 00:00:00 2001
From: dachengx <dx2227@columbia.edu>
Date: Sun, 23 Jul 2023 03:59:37 +0800
Subject: [PATCH 01/19] Add tests of blueice_extended_model

---
 alea/__init__.py                     |  2 +-
 tests/test_blueice_extended_model.py | 56 ++++++++++++++++++++++++++++
 tests/test_parameter.py              |  2 +
 tests/test_statistical_model.py      | 27 +++++++++++++-
 tests/test_template_source.py        |  3 ++
 5 files changed, 88 insertions(+), 2 deletions(-)
 create mode 100644 tests/test_blueice_extended_model.py
 create mode 100644 tests/test_parameter.py
 create mode 100644 tests/test_template_source.py

diff --git a/alea/__init__.py b/alea/__init__.py
index 99491ee8..95a14437 100644
--- a/alea/__init__.py
+++ b/alea/__init__.py
@@ -5,4 +5,4 @@
 from alea import template_source
 from alea import toymc_running
 from alea import utils
-from alea import runner
+# from alea import runner
diff --git a/tests/test_blueice_extended_model.py b/tests/test_blueice_extended_model.py
new file mode 100644
index 00000000..b7e0f864
--- /dev/null
+++ b/tests/test_blueice_extended_model.py
@@ -0,0 +1,56 @@
+import yaml
+import pkg_resources
+from unittest import TestCase
+
+from blueice.likelihood import LogLikelihoodSum
+from alea.blueice_extended_model import BlueiceExtendedModel, CustomAncillaryLikelihood
+
+
+class TestBlueiceExtendedModel(TestCase):
+    # self.assertIsInstance(json_as_text, str)
+    # self.assertEqual(json_as_dict, get_resource(self.json_file, fmt='json'))
+
+    def __init__(self, *args, **kwargs):
+        # TODO: super().__init__(*args, **kwargs)
+        super(TestBlueiceExtendedModel, self).__init__(*args, **kwargs)
+        # TODO: put the file path searching in utils.py
+        with open(pkg_resources.resource_filename(
+            'alea', 'examples/unbinned_wimp_statistical_model.yaml'), 'r') as f:
+            self.config = yaml.safe_load(f)
+        self.n_likelihood_terms = len(self.config['likelihood_config']['likelihood_terms'])
+        self.model = BlueiceExtendedModel(
+            parameter_definition=self.config['parameter_definition'],
+            likelihood_config=self.config['likelihood_config'],
+        )
+
+    def test_expectation_values(self):
+        expectation_values = self.model.get_expectation_values()
+        # TODO: assert expectation values after test template source
+        # self.assertEqual()
+
+    def test_generate_data(self):
+        data = self.model.generate_data()
+        self.assertEqual(
+            len(data), self.n_likelihood_terms + 2)
+        if not all(['source' in d.dtype.names for d in data[:-2]]):
+            raise ValueError('Data does not contain source information.')
+
+    def test_likelihood(self):
+        self.assertIsInstance(self.model._likelihood, LogLikelihoodSum)
+        self.assertIsInstance(self.model._likelihood.likelihood_list[-1], CustomAncillaryLikelihood)
+        self.assertEqual(
+            len(self.model._likelihood.likelihood_list),
+            self.n_likelihood_terms + 1)
+        self.model.data = self.model.generate_data()
+        self.model._likelihood()
+
+    def test_fit(self):
+        self.model.data = self.model.generate_data()
+        fit_result, max_llh = self.model.fit()
+        # TODO: check whether all parameters are in fit_result and whether fittable parameters are fitted
+
+
+class TestCustomAncillaryLikelihood(TestCase):
+    def test_ancillary_likelihood(self):
+        pass
+
diff --git a/tests/test_parameter.py b/tests/test_parameter.py
new file mode 100644
index 00000000..0daf895c
--- /dev/null
+++ b/tests/test_parameter.py
@@ -0,0 +1,2 @@
+from alea.parameters import Parameters
+
diff --git a/tests/test_statistical_model.py b/tests/test_statistical_model.py
index 08e9ea40..42d56d1a 100644
--- a/tests/test_statistical_model.py
+++ b/tests/test_statistical_model.py
@@ -1,2 +1,27 @@
+from alea.examples.gaussian_model import GaussianModel
+
+
 def test_gaussian_model():
-    pass
+    parameter_definition = {
+        'mu': {
+            'fit_guess': 0.0,
+            'fittable': True,
+            'nominal_value': 0.0
+        },
+        'sigma': {
+            'fit_guess': 1.0,
+            'fit_limits': [
+                0,
+                None
+            ],
+            'fittable': True,
+            'nominal_value': 1.0
+        }
+    }
+    simple_model = GaussianModel(
+        parameter_definition=parameter_definition)
+    simple_model.data = simple_model.generate_data(mu=0, sigma=2)
+    fit_result, max_llh = simple_model.fit()
+
+    toydata_file = 'simple_data.hdf5'
+    simple_model.store_data(toydata_file, [simple_model.data])
diff --git a/tests/test_template_source.py b/tests/test_template_source.py
new file mode 100644
index 00000000..43fe51c7
--- /dev/null
+++ b/tests/test_template_source.py
@@ -0,0 +1,3 @@
+from alea.template_source import TemplateSource
+
+

From 9fc8e43d50f1ef756e1500f4b082a1c18d94a816 Mon Sep 17 00:00:00 2001
From: dachengx <dx2227@columbia.edu>
Date: Sun, 23 Jul 2023 04:38:50 +0800
Subject: [PATCH 02/19] Add docstrings

---
 tests/test_blueice_extended_model.py | 17 +++++++++++------
 tests/test_parameter.py              |  1 -
 tests/test_statistical_model.py      | 13 +++++++------
 tests/test_template_source.py        |  2 --
 4 files changed, 18 insertions(+), 15 deletions(-)

diff --git a/tests/test_blueice_extended_model.py b/tests/test_blueice_extended_model.py
index b7e0f864..74360743 100644
--- a/tests/test_blueice_extended_model.py
+++ b/tests/test_blueice_extended_model.py
@@ -7,12 +7,11 @@
 
 
 class TestBlueiceExtendedModel(TestCase):
-    # self.assertIsInstance(json_as_text, str)
-    # self.assertEqual(json_as_dict, get_resource(self.json_file, fmt='json'))
+    """Test of the BlueiceExtendedModel class"""
 
     def __init__(self, *args, **kwargs):
-        # TODO: super().__init__(*args, **kwargs)
-        super(TestBlueiceExtendedModel, self).__init__(*args, **kwargs)
+        """Initialize the BlueiceExtendedModel class"""
+        super().__init__(*args, **kwargs)
         # TODO: put the file path searching in utils.py
         with open(pkg_resources.resource_filename(
             'alea', 'examples/unbinned_wimp_statistical_model.yaml'), 'r') as f:
@@ -24,11 +23,13 @@ def __init__(self, *args, **kwargs):
         )
 
     def test_expectation_values(self):
+        """Test of the expectation_values method"""
         expectation_values = self.model.get_expectation_values()
         # TODO: assert expectation values after test template source
         # self.assertEqual()
 
     def test_generate_data(self):
+        """Test of the generate_data method"""
         data = self.model.generate_data()
         self.assertEqual(
             len(data), self.n_likelihood_terms + 2)
@@ -36,6 +37,7 @@ def test_generate_data(self):
             raise ValueError('Data does not contain source information.')
 
     def test_likelihood(self):
+        """Test of the _likelihood attribute"""
         self.assertIsInstance(self.model._likelihood, LogLikelihoodSum)
         self.assertIsInstance(self.model._likelihood.likelihood_list[-1], CustomAncillaryLikelihood)
         self.assertEqual(
@@ -45,12 +47,15 @@ def test_likelihood(self):
         self.model._likelihood()
 
     def test_fit(self):
+        """Test of the fit method"""
         self.model.data = self.model.generate_data()
         fit_result, max_llh = self.model.fit()
-        # TODO: check whether all parameters are in fit_result and whether fittable parameters are fitted
+        # TODO: check whether all parameters are in fit_result
+        # and whether fittable parameters are fitted
 
 
 class TestCustomAncillaryLikelihood(TestCase):
+    """Test of the CustomAncillaryLikelihood class"""
     def test_ancillary_likelihood(self):
+        """Test of the ancillary_likelihood method"""
         pass
-
diff --git a/tests/test_parameter.py b/tests/test_parameter.py
index 0daf895c..1faf90de 100644
--- a/tests/test_parameter.py
+++ b/tests/test_parameter.py
@@ -1,2 +1 @@
 from alea.parameters import Parameters
-
diff --git a/tests/test_statistical_model.py b/tests/test_statistical_model.py
index 42d56d1a..82976f31 100644
--- a/tests/test_statistical_model.py
+++ b/tests/test_statistical_model.py
@@ -2,20 +2,21 @@
 
 
 def test_gaussian_model():
+    """Test of the GaussianModel class"""
     parameter_definition = {
         'mu': {
-            'fit_guess': 0.0,
+            'fit_guess': 0.,
             'fittable': True,
-            'nominal_value': 0.0
+            'nominal_value': 0.,
         },
         'sigma': {
-            'fit_guess': 1.0,
+            'fit_guess': 1.,
             'fit_limits': [
-                0,
-                None
+                0.,
+                None,
             ],
             'fittable': True,
-            'nominal_value': 1.0
+            'nominal_value': 1.,
         }
     }
     simple_model = GaussianModel(
diff --git a/tests/test_template_source.py b/tests/test_template_source.py
index 43fe51c7..d43d1df8 100644
--- a/tests/test_template_source.py
+++ b/tests/test_template_source.py
@@ -1,3 +1 @@
 from alea.template_source import TemplateSource
-
-

From ef1d7138c8aa7a7078b97d61c9bf441c1febbd1b Mon Sep 17 00:00:00 2001
From: dachengx <dx2227@columbia.edu>
Date: Sun, 23 Jul 2023 04:45:48 +0800
Subject: [PATCH 03/19] Add iminuit

---
 requirements.txt | 1 +
 1 file changed, 1 insertion(+)

diff --git a/requirements.txt b/requirements.txt
index bcc396a5..46a68d78 100644
--- a/requirements.txt
+++ b/requirements.txt
@@ -10,3 +10,4 @@ scipy
 setuptools
 tqdm
 mergedeep
+iminuit

From 7feabb4af0d2c2617693b2504288ca7e59717fd1 Mon Sep 17 00:00:00 2001
From: dachengx <dx2227@columbia.edu>
Date: Sun, 23 Jul 2023 04:53:56 +0800
Subject: [PATCH 04/19] Set data before get expectation values

---
 tests/test_blueice_extended_model.py | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/tests/test_blueice_extended_model.py b/tests/test_blueice_extended_model.py
index 74360743..62ed22ca 100644
--- a/tests/test_blueice_extended_model.py
+++ b/tests/test_blueice_extended_model.py
@@ -24,6 +24,7 @@ def __init__(self, *args, **kwargs):
 
     def test_expectation_values(self):
         """Test of the expectation_values method"""
+        self.model.data = self.model.generate_data()
         expectation_values = self.model.get_expectation_values()
         # TODO: assert expectation values after test template source
         # self.assertEqual()
@@ -56,6 +57,7 @@ def test_fit(self):
 
 class TestCustomAncillaryLikelihood(TestCase):
     """Test of the CustomAncillaryLikelihood class"""
+
     def test_ancillary_likelihood(self):
         """Test of the ancillary_likelihood method"""
         pass

From 33fe8ab50c1af1a85288df4c9bc8e2bb9f0ab2a8 Mon Sep 17 00:00:00 2001
From: dachengx <dx2227@columbia.edu>
Date: Tue, 25 Jul 2023 11:41:05 +0800
Subject: [PATCH 05/19] Change import path

---
 tests/test_statistical_model.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/tests/test_statistical_model.py b/tests/test_statistical_model.py
index 82976f31..6525e4d1 100644
--- a/tests/test_statistical_model.py
+++ b/tests/test_statistical_model.py
@@ -1,4 +1,4 @@
-from alea.examples.gaussian_model import GaussianModel
+from alea.models.gaussian_model import GaussianModel
 
 
 def test_gaussian_model():

From 59a0122e6dedaa3403088aa4bf71af1faef07052 Mon Sep 17 00:00:00 2001
From: dachengx <dx2227@columbia.edu>
Date: Tue, 25 Jul 2023 13:45:41 +0800
Subject: [PATCH 06/19] Add a fucntion get_file_path to get file path

---
 .../unbinned_wimp_statistical_model.yaml      |  2 +-
 alea/models/blueice_extended_model.py         |  2 +
 alea/utils.py                                 | 82 ++++++++++++++-----
 tests/test_blueice_extended_model.py          |  9 +-
 tests/test_statistical_model.py               |  2 +-
 5 files changed, 68 insertions(+), 29 deletions(-)

diff --git a/alea/model_configs/unbinned_wimp_statistical_model.yaml b/alea/model_configs/unbinned_wimp_statistical_model.yaml
index 2f69425d..c0665718 100644
--- a/alea/model_configs/unbinned_wimp_statistical_model.yaml
+++ b/alea/model_configs/unbinned_wimp_statistical_model.yaml
@@ -62,7 +62,7 @@ parameter_definition:
 
 likelihood_config:
   likelihood_weights: [1, 1, 1]
-  template_folder: alea/templates
+  template_folder: []
   likelihood_terms:
     # SR0
     - name: sr0
diff --git a/alea/models/blueice_extended_model.py b/alea/models/blueice_extended_model.py
index 6c5e511b..c2ab5f47 100644
--- a/alea/models/blueice_extended_model.py
+++ b/alea/models/blueice_extended_model.py
@@ -103,6 +103,8 @@ def _build_ll_from_config(self, likelihood_config: dict) -> "LogLikelihoodSum":
         # Iterate through each likelihood term in the configuration
         for config in likelihood_config["likelihood_terms"]:
             likelihood_object = locate(config["likelihood_type"])
+            if "template_folder" not in likelihood_config:
+                likelihood_config["template_folder"] = []
             if isinstance(likelihood_config["template_folder"], str):
                 template_folder_list = [likelihood_config["template_folder"]]
             elif isinstance(likelihood_config["template_folder"], list):
diff --git a/alea/utils.py b/alea/utils.py
index 834c9cd5..982cfee4 100644
--- a/alea/utils.py
+++ b/alea/utils.py
@@ -1,8 +1,10 @@
 import os
+import yaml
+import pkg_resources
 from pydoc import locate
+from warnings import warn
 
 import numpy as np
-import alea
 
 
 def get_analysis_space(analysis_space: dict) -> list:
@@ -31,29 +33,12 @@ def adapt_likelihood_config_for_blueice(
 
     Args:
         likelihood_config (dict): likelihood config dict
-        template_folder_list (list): list of possible base folders where
-            templates are located. If a folder starts with alea/,
-            the alea folder is used as base.
+        template_folder_list (list): list of possible base folders.
             Ordered by priority.
 
     Returns:
         dict: adapted likelihood config
     """
-    template_folder = None
-    for template_folder in template_folder_list:
-        # if template folder starts with alea: get location of alea
-        if template_folder.startswith("alea/"):
-            alea_dir = os.path.dirname(os.path.abspath(alea.__file__))
-            template_folder = os.path.join(alea_dir, template_folder.replace("alea/", ""))
-            # check if template folder exists
-            if not os.path.isdir(template_folder):
-                template_folder = None
-            else:
-                break
-
-    # raise error if no template folder is found
-    if template_folder is None:
-        raise FileNotFoundError("No template folder found. Please provide a valid template folder.")
 
     likelihood_config["analysis_space"] = get_analysis_space(
         likelihood_config["analysis_space"])
@@ -62,6 +47,61 @@ def adapt_likelihood_config_for_blueice(
         likelihood_config["default_source_class"])
 
     for source in likelihood_config["sources"]:
-        source["templatename"] = os.path.join(
-            template_folder, source["template_filename"])
+        source["templatename"] = get_file_path(
+            source["template_filename"], template_folder_list)
     return likelihood_config
+
+
+def load_yaml(file_name: str):
+    """Load data from yaml file."""
+    with open(get_file_path(file_name), 'r') as file:
+        data = yaml.safe_load(file)
+    return data
+
+
+def _get_abspath(file_name):
+    """Get the abspath of the file. Raise FileNotFoundError when not found in any subfolder"""
+    for sub_dir in ('model_configs', 'runner_configs', 'templates'):
+        p = os.path.join(_package_path(sub_dir), file_name)
+        if os.path.exists(p):
+            return p
+    raise FileNotFoundError(f'Cannot find {file_name}')
+
+
+def _package_path(sub_directory):
+    """Get the abs path of the requested sub folder"""
+    return pkg_resources.resource_filename('alea', f'{sub_directory}')
+
+
+def get_file_path(fname, folder_list=[]):
+    """Find the full path to the resource file
+    Try 5 methods in the following order
+
+    #. fname begin with '/', return absolute path
+    #. url_base begin with '/', return url_base + name
+    #. can get file from _get_abspath, return alea internal file path
+    #. can be found in local installed ntauxfiles, return ntauxfiles absolute path
+    #. can be downloaded from MongoDB, download and return cached path
+    """
+    # 1. From absolute path
+    # Usually Config.default is a absolute path
+    if fname.startswith('/'):
+        return fname
+
+    # 2. From local folder
+    # Use url_base as prefix
+    for folder in folder_list:
+        if folder.startswith('/'):
+            fpath = os.path.join(folder, fname)
+            if os.path.exists(fpath):
+                warn(f'Load {fname} successfully from {fpath}')
+                return fpath
+
+    # 3. From alea internal files
+    try:
+        return _get_abspath(fname)
+    except FileNotFoundError:
+        pass
+
+    # raise error when can not find corresponding file
+    raise RuntimeError(f'Can not find {fname}, please check your file system')
diff --git a/tests/test_blueice_extended_model.py b/tests/test_blueice_extended_model.py
index 62ed22ca..f3359367 100644
--- a/tests/test_blueice_extended_model.py
+++ b/tests/test_blueice_extended_model.py
@@ -1,9 +1,9 @@
-import yaml
 import pkg_resources
 from unittest import TestCase
 
 from blueice.likelihood import LogLikelihoodSum
-from alea.blueice_extended_model import BlueiceExtendedModel, CustomAncillaryLikelihood
+from alea.utils import load_yaml
+from alea.models import BlueiceExtendedModel, CustomAncillaryLikelihood
 
 
 class TestBlueiceExtendedModel(TestCase):
@@ -12,10 +12,7 @@ class TestBlueiceExtendedModel(TestCase):
     def __init__(self, *args, **kwargs):
         """Initialize the BlueiceExtendedModel class"""
         super().__init__(*args, **kwargs)
-        # TODO: put the file path searching in utils.py
-        with open(pkg_resources.resource_filename(
-            'alea', 'examples/unbinned_wimp_statistical_model.yaml'), 'r') as f:
-            self.config = yaml.safe_load(f)
+        self.config = load_yaml('unbinned_wimp_statistical_model.yaml')
         self.n_likelihood_terms = len(self.config['likelihood_config']['likelihood_terms'])
         self.model = BlueiceExtendedModel(
             parameter_definition=self.config['parameter_definition'],
diff --git a/tests/test_statistical_model.py b/tests/test_statistical_model.py
index 6525e4d1..940389de 100644
--- a/tests/test_statistical_model.py
+++ b/tests/test_statistical_model.py
@@ -1,4 +1,4 @@
-from alea.models.gaussian_model import GaussianModel
+from alea.models import GaussianModel
 
 
 def test_gaussian_model():

From ca81ac2bd635f59e17cf6cdfd6b26b8e19add9ba Mon Sep 17 00:00:00 2001
From: dachengx <dx2227@columbia.edu>
Date: Tue, 25 Jul 2023 14:53:29 +0800
Subject: [PATCH 07/19] Make Parameters deepcopyable

---
 alea/model.py                         |  9 +++++++-
 alea/models/blueice_extended_model.py | 24 ++++++++++++-------
 alea/parameters.py                    | 11 +++++----
 tests/test_blueice_extended_model.py  | 15 ++++++++++--
 tests/test_gaussian_model.py          | 33 +++++++++++++++++++++++++++
 tests/test_parameter.py               | 19 +++++++++++++++
 tests/test_statistical_model.py       | 28 -----------------------
 tests/test_template_source.py         |  8 +++++++
 8 files changed, 103 insertions(+), 44 deletions(-)
 create mode 100644 tests/test_gaussian_model.py
 delete mode 100644 tests/test_statistical_model.py

diff --git a/alea/model.py b/alea/model.py
index 82a327e9..b7fe59b7 100644
--- a/alea/model.py
+++ b/alea/model.py
@@ -7,6 +7,7 @@
 from scipy.optimize import brentq
 from iminuit import Minuit
 from iminuit.util import make_func_code
+from blueice.likelihood import _needs_data
 from inference_interface import toydata_to_file
 
 from alea.parameters import Parameters
@@ -58,7 +59,10 @@ def __init__(
             confidence_interval_kind: str = "central",  # one of central, upper, lower
             confidence_interval_threshold: Callable[[float], float] = None,
         ):
-        self._data = data
+        # following https://github.com/JelleAalbers/blueice/blob/7c10222a13227e78dc7224b1a7e56ff91e4a8043/blueice/likelihood.py#L97
+        self.is_data_set = False
+        if data is not None:
+            self.data = data
         self._confidence_level = confidence_level
         self._confidence_interval_kind = confidence_interval_kind
         self.confidence_interval_threshold = confidence_interval_threshold
@@ -93,6 +97,7 @@ def _generate_data(self, **kwargs):
             "You must write a data-generation method (_generate_data) for your statistical model"
             " or use a subclass where it is written for you")
 
+    @_needs_data
     def ll(self, **kwargs) -> float:
         """
         Likelihod function, returns the loglikelihood for the given parameters.
@@ -143,6 +148,7 @@ def data(self, data):
         representing the data-sets of one or more likelihood terms.
         """
         self._data = data
+        self.is_data_set = True
 
     def store_data(
             self, file_name, data_list, data_name_list=None, metadata = None):
@@ -205,6 +211,7 @@ def cost(args):
 
         return cost
 
+    @_needs_data
     def fit(self, verbose=False, **kwargs) -> Tuple[dict, float]:
         """
         Fit the model to the data by maximizing the likelihood
diff --git a/alea/models/blueice_extended_model.py b/alea/models/blueice_extended_model.py
index c2ab5f47..1427cb32 100644
--- a/alea/models/blueice_extended_model.py
+++ b/alea/models/blueice_extended_model.py
@@ -1,5 +1,6 @@
-from pydoc import locate  # to lookup likelihood class
 from typing import List
+from copy import deepcopy
+from pydoc import locate
 
 import yaml
 import numpy as np
@@ -7,9 +8,9 @@
 from blueice.likelihood import LogAncillaryLikelihood, LogLikelihoodSum
 
 from alea.model import StatisticalModel
+from alea.parameters import Parameters
 from alea.simulators import BlueiceDataGenerator
 from alea.utils import adapt_likelihood_config_for_blueice
-from alea.parameters import Parameters
 
 
 class BlueiceExtendedModel(StatisticalModel):
@@ -32,7 +33,9 @@ def __init__(self, parameter_definition: dict, likelihood_config: dict):
             likelihood_config (dict): A dictionary defining the likelihood.
         """
         super().__init__(parameter_definition=parameter_definition)
-        self._likelihood = self._build_ll_from_config(likelihood_config)
+        # deepcopy likelihood_config to prevent it to be
+        # changed by adapt_likelihood_config_for_blueice
+        self._likelihood = self._build_ll_from_config(deepcopy(likelihood_config))
         self.likelihood_names = [t["name"] for t in likelihood_config["likelihood_terms"]]
         self.likelihood_names.append("ancillary_likelihood")
         self.data_generators = self._build_data_generators()
@@ -72,6 +75,7 @@ def data(self, data: list):
             ll_term.set_data(d)
 
         self._data = data
+        self.is_data_set = True
 
     def get_expectation_values(self, **kwargs) -> dict:
         """
@@ -79,15 +83,19 @@ def get_expectation_values(self, **kwargs) -> dict:
         given a number of named parameters (kwargs)
         """
         ret = dict()
-        # ancillary likelihood does not contribute
-        for ll in self._likelihood.likelihood_list[:-1]:
 
-            ll_pars = list(ll.rate_parameters.keys()) + list(ll.shape_parameters.keys())
+        # calling ll need data to be set
+        self_copy = deepcopy(self)
+        self_copy.data = self_copy.generate_data()
+
+        # ancillary likelihood does not contribute
+        for ll_term in self_copy._likelihood.likelihood_list[:-1]:
+            ll_pars = list(ll_term.rate_parameters.keys()) + list(ll_term.shape_parameters.keys())
             ll_pars += ["livetime_days"]
             call_args = {k: i for k, i in kwargs.items() if k in ll_pars}
 
-            mus = ll(full_output=True, **call_args)[1]
-            for n, mu in zip(ll.source_name_list, mus):
+            mus = ll_term(full_output=True, **call_args)[1]
+            for n, mu in zip(ll_term.source_name_list, mus):
                 ret[n] = ret.get(n, 0) + mu
         return ret
 
diff --git a/alea/parameters.py b/alea/parameters.py
index 589f5fc7..e7f26da0 100644
--- a/alea/parameters.py
+++ b/alea/parameters.py
@@ -274,8 +274,9 @@ def __call__(
                 values[name] = new_val if new_val is not None else param.nominal_value
         if any(i is None for k, i in values.items()):
             emptypars = ", ".join([k for k, i in values.items() if i is None])
-            raise AssertionError("All parameters must be set explicitly, or have a nominal value,"
-                                 " encountered for: " + emptypars)
+            raise AssertionError(
+                "All parameters must be set explicitly, or have a nominal value,"
+                " encountered for: " + emptypars)
         return values
 
     def __getattr__(self, name: str) -> Parameter:
@@ -291,9 +292,9 @@ def __getattr__(self, name: str) -> Parameter:
         Raises:
             AttributeError: If the attribute is not found.
         """
-        if name in self.parameters:
-            return self.parameters[name]
-        else:
+        try:
+            return super().__getattribute__('parameters')[name]
+        except KeyError:
             raise AttributeError(f"Attribute '{name}' not found.")
 
     def __getitem__(self, name: str) -> Parameter:
diff --git a/tests/test_blueice_extended_model.py b/tests/test_blueice_extended_model.py
index f3359367..ce9789a6 100644
--- a/tests/test_blueice_extended_model.py
+++ b/tests/test_blueice_extended_model.py
@@ -1,4 +1,3 @@
-import pkg_resources
 from unittest import TestCase
 
 from blueice.likelihood import LogLikelihoodSum
@@ -14,6 +13,9 @@ def __init__(self, *args, **kwargs):
         super().__init__(*args, **kwargs)
         self.config = load_yaml('unbinned_wimp_statistical_model.yaml')
         self.n_likelihood_terms = len(self.config['likelihood_config']['likelihood_terms'])
+        self.set_new_model()
+
+    def set_new_model(self):
         self.model = BlueiceExtendedModel(
             parameter_definition=self.config['parameter_definition'],
             likelihood_config=self.config['likelihood_config'],
@@ -21,8 +23,16 @@ def __init__(self, *args, **kwargs):
 
     def test_expectation_values(self):
         """Test of the expectation_values method"""
-        self.model.data = self.model.generate_data()
+        self.set_new_model()
         expectation_values = self.model.get_expectation_values()
+
+        # should avoid accidentally set data
+        is_data_set = False
+        for ll_term in self.model._likelihood.likelihood_list[:-1]:
+            is_data_set |= ll_term.is_data_set
+        if is_data_set:
+            raise ValueError('Data should not be set after get_expectation_values.')
+
         # TODO: assert expectation values after test template source
         # self.assertEqual()
 
@@ -57,4 +67,5 @@ class TestCustomAncillaryLikelihood(TestCase):
 
     def test_ancillary_likelihood(self):
         """Test of the ancillary_likelihood method"""
+        # TODO:
         pass
diff --git a/tests/test_gaussian_model.py b/tests/test_gaussian_model.py
new file mode 100644
index 00000000..c668587a
--- /dev/null
+++ b/tests/test_gaussian_model.py
@@ -0,0 +1,33 @@
+from unittest import TestCase
+
+from alea.models import GaussianModel
+
+
+class TestGaussianModel(TestCase):
+    """Test of the Parameters class"""
+
+    def test_gaussian_model(self):
+        """Test of generate_data and fit method of the GaussianModel class"""
+        parameter_definition = {
+            'mu': {
+                'fit_guess': 0.,
+                'fittable': True,
+                'nominal_value': 0.,
+            },
+            'sigma': {
+                'fit_guess': 1.,
+                'fit_limits': [
+                    0.,
+                    None,
+                ],
+                'fittable': True,
+                'nominal_value': 1.,
+            }
+        }
+        simple_model = GaussianModel(
+            parameter_definition=parameter_definition)
+        simple_model.data = simple_model.generate_data(mu=0, sigma=2)
+        fit_result, max_llh = simple_model.fit()
+
+        toydata_file = 'simple_data.hdf5'
+        simple_model.store_data(toydata_file, [simple_model.data])
diff --git a/tests/test_parameter.py b/tests/test_parameter.py
index 1faf90de..876788f4 100644
--- a/tests/test_parameter.py
+++ b/tests/test_parameter.py
@@ -1 +1,20 @@
+from copy import deepcopy
+from unittest import TestCase
+
+from alea.utils import load_yaml
 from alea.parameters import Parameters
+
+
+class TestParameters(TestCase):
+    """Test of the Parameters class"""
+
+    def __init__(self, *args, **kwargs):
+        """Initialize the BlueiceExtendedModel class"""
+        super().__init__(*args, **kwargs)
+        self.config = load_yaml('unbinned_wimp_statistical_model.yaml')
+        self.parameters = Parameters.from_config(self.config['parameter_definition'])
+
+    def test_deep_copyable(self):
+        """Test of whether Parameters instance can be deepcopied"""
+        if deepcopy(self.parameters) != self.parameters:
+            raise ValueError('Parameters instance cannot be correctly deepcopied.')
diff --git a/tests/test_statistical_model.py b/tests/test_statistical_model.py
deleted file mode 100644
index 940389de..00000000
--- a/tests/test_statistical_model.py
+++ /dev/null
@@ -1,28 +0,0 @@
-from alea.models import GaussianModel
-
-
-def test_gaussian_model():
-    """Test of the GaussianModel class"""
-    parameter_definition = {
-        'mu': {
-            'fit_guess': 0.,
-            'fittable': True,
-            'nominal_value': 0.,
-        },
-        'sigma': {
-            'fit_guess': 1.,
-            'fit_limits': [
-                0.,
-                None,
-            ],
-            'fittable': True,
-            'nominal_value': 1.,
-        }
-    }
-    simple_model = GaussianModel(
-        parameter_definition=parameter_definition)
-    simple_model.data = simple_model.generate_data(mu=0, sigma=2)
-    fit_result, max_llh = simple_model.fit()
-
-    toydata_file = 'simple_data.hdf5'
-    simple_model.store_data(toydata_file, [simple_model.data])
diff --git a/tests/test_template_source.py b/tests/test_template_source.py
index d43d1df8..f57cb30b 100644
--- a/tests/test_template_source.py
+++ b/tests/test_template_source.py
@@ -1 +1,9 @@
+from unittest import TestCase
+
 from alea.template_source import TemplateSource
+
+
+class TestTemplateSource(TestCase):
+    """Test of the TemplateSource class"""
+    # TODO:
+    pass

From 82e858d57f88c6b2a2d4bf1975e8787a2694db96 Mon Sep 17 00:00:00 2001
From: dachengx <dx2227@columbia.edu>
Date: Tue, 25 Jul 2023 14:58:35 +0800
Subject: [PATCH 08/19] Happier code style

---
 alea/model.py                        | 3 ++-
 alea/utils.py                        | 4 +++-
 tests/test_blueice_extended_model.py | 1 +
 tests/test_gaussian_model.py         | 2 +-
 tests/test_template_source.py        | 6 ++++--
 5 files changed, 11 insertions(+), 5 deletions(-)

diff --git a/alea/model.py b/alea/model.py
index b7fe59b7..fbbc630d 100644
--- a/alea/model.py
+++ b/alea/model.py
@@ -59,7 +59,8 @@ def __init__(
             confidence_interval_kind: str = "central",  # one of central, upper, lower
             confidence_interval_threshold: Callable[[float], float] = None,
         ):
-        # following https://github.com/JelleAalbers/blueice/blob/7c10222a13227e78dc7224b1a7e56ff91e4a8043/blueice/likelihood.py#L97
+        # following https://github.com/JelleAalbers/blueice/blob/
+        # 7c10222a13227e78dc7224b1a7e56ff91e4a8043/blueice/likelihood.py#L97
         self.is_data_set = False
         if data is not None:
             self.data = data
diff --git a/alea/utils.py b/alea/utils.py
index 982cfee4..a6485fd6 100644
--- a/alea/utils.py
+++ b/alea/utils.py
@@ -73,7 +73,7 @@ def _package_path(sub_directory):
     return pkg_resources.resource_filename('alea', f'{sub_directory}')
 
 
-def get_file_path(fname, folder_list=[]):
+def get_file_path(fname, folder_list=None):
     """Find the full path to the resource file
     Try 5 methods in the following order
 
@@ -83,6 +83,8 @@ def get_file_path(fname, folder_list=[]):
     #. can be found in local installed ntauxfiles, return ntauxfiles absolute path
     #. can be downloaded from MongoDB, download and return cached path
     """
+    if folder_list is None:
+        folder_list = []
     # 1. From absolute path
     # Usually Config.default is a absolute path
     if fname.startswith('/'):
diff --git a/tests/test_blueice_extended_model.py b/tests/test_blueice_extended_model.py
index ce9789a6..f388a78e 100644
--- a/tests/test_blueice_extended_model.py
+++ b/tests/test_blueice_extended_model.py
@@ -16,6 +16,7 @@ def __init__(self, *args, **kwargs):
         self.set_new_model()
 
     def set_new_model(self):
+        """Set a new BlueiceExtendedModel instance"""
         self.model = BlueiceExtendedModel(
             parameter_definition=self.config['parameter_definition'],
             likelihood_config=self.config['likelihood_config'],
diff --git a/tests/test_gaussian_model.py b/tests/test_gaussian_model.py
index c668587a..5caf0d7f 100644
--- a/tests/test_gaussian_model.py
+++ b/tests/test_gaussian_model.py
@@ -22,7 +22,7 @@ def test_gaussian_model(self):
                 ],
                 'fittable': True,
                 'nominal_value': 1.,
-            }
+            },
         }
         simple_model = GaussianModel(
             parameter_definition=parameter_definition)
diff --git a/tests/test_template_source.py b/tests/test_template_source.py
index f57cb30b..1088b233 100644
--- a/tests/test_template_source.py
+++ b/tests/test_template_source.py
@@ -5,5 +5,7 @@
 
 class TestTemplateSource(TestCase):
     """Test of the TemplateSource class"""
-    # TODO:
-    pass
+
+    def test_load_templates(self):
+        # TODO:
+        pass

From 166033e9a3157d40bc41f009d2e919a6c8348245 Mon Sep 17 00:00:00 2001
From: dachengx <dx2227@columbia.edu>
Date: Tue, 25 Jul 2023 17:20:53 +0800
Subject: [PATCH 09/19] Move template_folder_list outside the loop

---
 alea/__init__.py                      |  3 +++
 alea/models/blueice_extended_model.py | 20 +++++++-------
 alea/template_source.py               | 39 ++++++++++++++-------------
 alea/utils.py                         | 15 ++++++-----
 tests/test_template_source.py         |  3 ++-
 5 files changed, 44 insertions(+), 36 deletions(-)

diff --git a/alea/__init__.py b/alea/__init__.py
index 797c9f55..f008bb17 100644
--- a/alea/__init__.py
+++ b/alea/__init__.py
@@ -13,3 +13,6 @@
 
 from . import simulators
 from .simulators import *
+
+from . import template_source
+from .template_source import *
diff --git a/alea/models/blueice_extended_model.py b/alea/models/blueice_extended_model.py
index 1427cb32..9e66eec6 100644
--- a/alea/models/blueice_extended_model.py
+++ b/alea/models/blueice_extended_model.py
@@ -108,18 +108,19 @@ def _build_ll_from_config(self, likelihood_config: dict) -> "LogLikelihoodSum":
         """
         lls = []
 
+        if "template_folder" not in likelihood_config:
+            likelihood_config["template_folder"] = []
+        if isinstance(likelihood_config["template_folder"], str):
+            template_folder_list = [likelihood_config["template_folder"]]
+        elif isinstance(likelihood_config["template_folder"], list):
+            template_folder_list = likelihood_config["template_folder"]
+        else:
+            raise ValueError(
+                "template_folder must be either a string or a list of strings.")
+
         # Iterate through each likelihood term in the configuration
         for config in likelihood_config["likelihood_terms"]:
             likelihood_object = locate(config["likelihood_type"])
-            if "template_folder" not in likelihood_config:
-                likelihood_config["template_folder"] = []
-            if isinstance(likelihood_config["template_folder"], str):
-                template_folder_list = [likelihood_config["template_folder"]]
-            elif isinstance(likelihood_config["template_folder"], list):
-                template_folder_list = likelihood_config["template_folder"]
-            else:
-                raise ValueError(
-                    "template_folder must be either a string or a list of strings.")
 
             blueice_config = adapt_likelihood_config_for_blueice(
                 config, template_folder_list)
@@ -143,7 +144,6 @@ def _build_ll_from_config(self, likelihood_config: dict) -> "LogLikelihoodSum":
             ll = likelihood_object(blueice_config)
 
             for source in config["sources"]:
-
                 # Set rate parameters
                 rate_parameters = [
                     p for p in source["parameters"] if self.parameters[p].ptype == "rate"]
diff --git a/alea/template_source.py b/alea/template_source.py
index d5c31f34..dd4480c0 100644
--- a/alea/template_source.py
+++ b/alea/template_source.py
@@ -24,6 +24,7 @@ class TemplateSource(blueice.HistogramPdfSource):
     :param log10_bins: List of axis numbers.
     If True, bin edges on this axis in the root file are log10() of the actual bin edges.
     """
+
     def build_histogram(self):
         format_dict = {
             k: self.config[k]
@@ -71,7 +72,6 @@ def build_histogram(self):
                                 stop=slice_axis_limits[1])
                     logging.debug(f"Normalization after slicing: {h.n}.")
 
-
             if collapse_axis is not None:
                 if collapse_slices is None:
                     raise ValueError(
@@ -94,9 +94,8 @@ def build_histogram(self):
                     self.config['analysis_space']):
                 expected_bin_edges = np.array(expected_bin_edges)
                 seen_bin_edges = h.bin_edges[axis_i]
-                if len(
-                        self.config['analysis_space']
-                ) == 1:  # If 1D, hist1d returns bin_edges straight, not as list
+                # If 1D, hist1d returns bin_edges straight, not as list
+                if len(self.config['analysis_space']) == 1:
                     seen_bin_edges = h.bin_edges
                 logging.debug("axis_i: " + str(axis_i))
                 logging.debug("expected_bin_edges: " + str(expected_bin_edges))
@@ -175,6 +174,7 @@ class CombinedSource(blueice.HistogramPdfSource):
     Must be 1 shorter than histnames, templatenames
     :param histogram_parameters: names of parameters that should be put in the hdf5/histogram names,
     """
+
     def build_histogram(self):
         weight_names = self.config.get("weight_names")
         weights = [
@@ -355,20 +355,6 @@ def simulate(self, n_events):
         return ret
 
 
-def get_json_spectrum(fn):
-    """
-    Translates bbf-style JSON files to spectra.
-    units are keV and /kev*day*kg
-    """
-    contents = json.load(open(fn, "r"))
-    logging.debug(contents["description"])
-    esyst = contents["coordinate_system"][0][1]
-    ret = interp1d(
-        np.linspace(*esyst), contents["map"],
-        bounds_error=False, fill_value=0.)
-    return ret
-
-
 class SpectrumTemplateSource(blueice.HistogramPdfSource):
     """
     :param spectrum_name: name of bbf json-like spectrum _OR_ function that can be called
@@ -376,6 +362,21 @@ class SpectrumTemplateSource(blueice.HistogramPdfSource):
     :param histname: histogram name
     :param named_parameters: list of config settings to pass to .format on histname and filename
     """
+
+    @staticmethod
+    def _get_json_spectrum(fn):
+        """
+        Translates bbf-style JSON files to spectra.
+        units are keV and /kev*day*kg
+        """
+        contents = json.load(open(fn, "r"))
+        logging.debug(contents["description"])
+        esyst = contents["coordinate_system"][0][1]
+        ret = interp1d(
+            np.linspace(*esyst), contents["map"],
+            bounds_error=False, fill_value=0.)
+        return ret
+
     def build_histogram(self):
         logging.debug("building a hist")
         format_dict = {
@@ -387,7 +388,7 @@ def build_histogram(self):
 
         spectrum = self.config["spectrum"]
         if type(spectrum) is str:
-            spectrum = get_json_spectrum(spectrum.format(**format_dict))
+            spectrum = self._get_json_spectrum(spectrum.format(**format_dict))
 
         slice_args = self.config.get("slice_args", {})
         if type(slice_args) is dict:
diff --git a/alea/utils.py b/alea/utils.py
index a6485fd6..631c87f0 100644
--- a/alea/utils.py
+++ b/alea/utils.py
@@ -1,6 +1,7 @@
 import os
 import yaml
 import pkg_resources
+from copy import deepcopy
 from pydoc import locate
 from warnings import warn
 
@@ -40,16 +41,18 @@ def adapt_likelihood_config_for_blueice(
         dict: adapted likelihood config
     """
 
-    likelihood_config["analysis_space"] = get_analysis_space(
-        likelihood_config["analysis_space"])
+    likelihood_config_copy = deepcopy(likelihood_config)
 
-    likelihood_config["default_source_class"] = locate(
-        likelihood_config["default_source_class"])
+    likelihood_config_copy["analysis_space"] = get_analysis_space(
+        likelihood_config_copy["analysis_space"])
 
-    for source in likelihood_config["sources"]:
+    likelihood_config_copy["default_source_class"] = locate(
+        likelihood_config_copy["default_source_class"])
+
+    for source in likelihood_config_copy["sources"]:
         source["templatename"] = get_file_path(
             source["template_filename"], template_folder_list)
-    return likelihood_config
+    return likelihood_config_copy
 
 
 def load_yaml(file_name: str):
diff --git a/tests/test_template_source.py b/tests/test_template_source.py
index 1088b233..36c4f4fd 100644
--- a/tests/test_template_source.py
+++ b/tests/test_template_source.py
@@ -7,5 +7,6 @@ class TestTemplateSource(TestCase):
     """Test of the TemplateSource class"""
 
     def test_load_templates(self):
-        # TODO:
+        # TODO: not sure whether we want to test TemplateSource in alea
+        # not sure TemplateSource.build_histogram and TemplateSource.simulate are called
         pass

From e769327d8d0a2ef81a8dfc625407194519213962 Mon Sep 17 00:00:00 2001
From: dachengx <dx2227@columbia.edu>
Date: Tue, 25 Jul 2023 17:43:16 +0800
Subject: [PATCH 10/19] Raise error when initialize StatisticalModel directly

---
 alea/model.py                   |  7 +++++++
 tests/test_statistical_model.py | 13 +++++++++++++
 2 files changed, 20 insertions(+)
 create mode 100644 tests/test_statistical_model.py

diff --git a/alea/model.py b/alea/model.py
index fbbc630d..8628df7e 100644
--- a/alea/model.py
+++ b/alea/model.py
@@ -59,6 +59,13 @@ def __init__(
             confidence_interval_kind: str = "central",  # one of central, upper, lower
             confidence_interval_threshold: Callable[[float], float] = None,
         ):
+        """Initialize a statistical model"""
+        if type(self) == StatisticalModel:
+            raise RuntimeError(
+                "You cannot instantiate the StatisticalModel class directly, "
+                "you must use a subclass where the likelihood function and data generation "
+                "method are implemented")
+
         # following https://github.com/JelleAalbers/blueice/blob/
         # 7c10222a13227e78dc7224b1a7e56ff91e4a8043/blueice/likelihood.py#L97
         self.is_data_set = False
diff --git a/tests/test_statistical_model.py b/tests/test_statistical_model.py
new file mode 100644
index 00000000..3dd09526
--- /dev/null
+++ b/tests/test_statistical_model.py
@@ -0,0 +1,13 @@
+from alea import StatisticalModel
+
+def test_gaussian_model():
+    try:
+        error_raised = True
+        StatisticalModel()
+        error_raised = False
+    except Exception:
+        print('Error correctly raised when directly instantiating StatisticalModel')
+    else:
+        if not error_raised:
+            raise RuntimeError(
+                'Should raise error when directly instantiating StatisticalModel')

From 3c27ec74f3f7c88db05d11dcfefe8ac36ff36a96 Mon Sep 17 00:00:00 2001
From: dachengx <dx2227@columbia.edu>
Date: Tue, 25 Jul 2023 17:52:06 +0800
Subject: [PATCH 11/19] Do not need to be too cautious because
 166033e9a3157d40bc41f009d2e919a6c8348245

---
 alea/models/blueice_extended_model.py | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)

diff --git a/alea/models/blueice_extended_model.py b/alea/models/blueice_extended_model.py
index 9e66eec6..49db7203 100644
--- a/alea/models/blueice_extended_model.py
+++ b/alea/models/blueice_extended_model.py
@@ -33,9 +33,7 @@ def __init__(self, parameter_definition: dict, likelihood_config: dict):
             likelihood_config (dict): A dictionary defining the likelihood.
         """
         super().__init__(parameter_definition=parameter_definition)
-        # deepcopy likelihood_config to prevent it to be
-        # changed by adapt_likelihood_config_for_blueice
-        self._likelihood = self._build_ll_from_config(deepcopy(likelihood_config))
+        self._likelihood = self._build_ll_from_config(likelihood_config)
         self.likelihood_names = [t["name"] for t in likelihood_config["likelihood_terms"]]
         self.likelihood_names.append("ancillary_likelihood")
         self.data_generators = self._build_data_generators()

From a1654dc66d0c9b7fb3862c41442e5b8e8576029d Mon Sep 17 00:00:00 2001
From: dachengx <dx2227@columbia.edu>
Date: Tue, 25 Jul 2023 18:30:15 +0800
Subject: [PATCH 12/19] This is why we want to accelerate the unittest

---
 alea/models/blueice_extended_model.py | 7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)

diff --git a/alea/models/blueice_extended_model.py b/alea/models/blueice_extended_model.py
index 4e54cb09..ddfe47f4 100644
--- a/alea/models/blueice_extended_model.py
+++ b/alea/models/blueice_extended_model.py
@@ -6,7 +6,7 @@
 import numpy as np
 import scipy.stats as stats
 from blueice.likelihood import LogAncillaryLikelihood, LogLikelihoodSum
-from inference_interface import dict_to_structured_array
+from inference_interface import dict_to_structured_array, structured_array_to_dict
 
 from alea.model import StatisticalModel
 from alea.parameters import Parameters
@@ -291,10 +291,11 @@ def set_data(self, d: np.array):
             d (np.array): Data of ancillary measurements, stored as numpy array
         """
         # This results in shifted constraint terms.
-        if set(d.keys()) != set(self.parameters.names):
+        d_dict = structured_array_to_dict(d)
+        if set(d_dict.keys()) != set(self.parameters.names):
             raise ValueError(
                 "The data dict must contain all parameters as keys in CustomAncillaryLikelihood.")
-        self.constraint_functions = self._get_constraint_functions(**d)
+        self.constraint_functions = self._get_constraint_functions(**d_dict)
 
     def ancillary_likelihood_sum(self, evaluate_at: dict) -> float:
         """Return the sum of all constraint terms.

From 37f00ff0aca9f9797c46d4a94d9a21e19e65f57e Mon Sep 17 00:00:00 2001
From: dachengx <dx2227@columbia.edu>
Date: Tue, 25 Jul 2023 23:08:29 +0800
Subject: [PATCH 13/19] Directly call ll of model

---
 tests/test_blueice_extended_model.py               | 6 ++----
 tests/{test_statistical_model.py => test_model.py} | 2 +-
 2 files changed, 3 insertions(+), 5 deletions(-)
 rename tests/{test_statistical_model.py => test_model.py} (92%)

diff --git a/tests/test_blueice_extended_model.py b/tests/test_blueice_extended_model.py
index f388a78e..a0897478 100644
--- a/tests/test_blueice_extended_model.py
+++ b/tests/test_blueice_extended_model.py
@@ -33,9 +33,7 @@ def test_expectation_values(self):
             is_data_set |= ll_term.is_data_set
         if is_data_set:
             raise ValueError('Data should not be set after get_expectation_values.')
-
-        # TODO: assert expectation values after test template source
-        # self.assertEqual()
+        # TODO: assert expectation values after manually initialize template source
 
     def test_generate_data(self):
         """Test of the generate_data method"""
@@ -53,7 +51,7 @@ def test_likelihood(self):
             len(self.model._likelihood.likelihood_list),
             self.n_likelihood_terms + 1)
         self.model.data = self.model.generate_data()
-        self.model._likelihood()
+        self.model.ll()
 
     def test_fit(self):
         """Test of the fit method"""
diff --git a/tests/test_statistical_model.py b/tests/test_model.py
similarity index 92%
rename from tests/test_statistical_model.py
rename to tests/test_model.py
index 3dd09526..5d99488b 100644
--- a/tests/test_statistical_model.py
+++ b/tests/test_model.py
@@ -1,6 +1,6 @@
 from alea import StatisticalModel
 
-def test_gaussian_model():
+def test_statistical_model():
     try:
         error_raised = True
         StatisticalModel()

From 52733344461b3d7f96fc78039fb01852c62135cf Mon Sep 17 00:00:00 2001
From: Knut Dundas Moraa <knutdundasmoraa@gmail.com>
Date: Tue, 25 Jul 2023 14:52:26 -0400
Subject: [PATCH 14/19] some more fitting tests

---
 tests/test_gaussian_model.py | 44 +++++++++++++++++++++++++++++++-----
 1 file changed, 38 insertions(+), 6 deletions(-)

diff --git a/tests/test_gaussian_model.py b/tests/test_gaussian_model.py
index 5caf0d7f..f2412a25 100644
--- a/tests/test_gaussian_model.py
+++ b/tests/test_gaussian_model.py
@@ -1,13 +1,24 @@
 from unittest import TestCase
 
+from os import remove
+
+import inference_interface
+
+import numpy as np
+
+import scipy.stats as sps
+
 from alea.models import GaussianModel
 
 
 class TestGaussianModel(TestCase):
     """Test of the Parameters class"""
 
-    def test_gaussian_model(self):
-        """Test of generate_data and fit method of the GaussianModel class"""
+    @classmethod
+    def setUp(cls):
+        """
+        Initialise the GaussianModel
+        """
         parameter_definition = {
             'mu': {
                 'fit_guess': 0.,
@@ -24,10 +35,31 @@ def test_gaussian_model(self):
                 'nominal_value': 1.,
             },
         }
-        simple_model = GaussianModel(
+        cls.simple_model = GaussianModel(
             parameter_definition=parameter_definition)
-        simple_model.data = simple_model.generate_data(mu=0, sigma=2)
-        fit_result, max_llh = simple_model.fit()
 
+    def test_data_generation(self):
+        """
+        Test of generate_data and fit method of the GaussianModel class
+        """
+        # test data generation:
+        self.simple_model.data = self.simple_model.generate_data(mu=0, sigma=2)
+
+    def test_data_storage(self):
+        # test data store+load:
         toydata_file = 'simple_data.hdf5'
-        simple_model.store_data(toydata_file, [simple_model.data])
+        self.simple_model.data = self.simple_model.generate_data(mu=0, sigma=2)
+        self.simple_model.store_data(toydata_file, [self.simple_model.data])
+        stored_data = inference_interface.toydata_from_file('simple_data.hdf5')
+        assert self.simple_model.data == stored_data[0] , "Stored data disagrees with data!"
+
+        remove("simple_data.hdf5")
+
+    def test_fit(self):
+        # test fitting:
+        self.simple_model.data = self.simple_model.generate_data(mu=0, sigma=2)
+        hat_meas = self.simple_model.data[0]["hat_mu"]
+        best_fit, lf = self.simple_model.fit(sigma=2)
+        hat_fit = best_fit["mu"]
+        np.testing.assert_almost_equal(hat_meas, hat_fit), "best-fit does not agree"
+        np.testing.assert_almost_equal(lf, sps.norm(hat_fit, 2).logpdf(hat_meas)) , "likelihood function disagrees"
\ No newline at end of file

From 264e663ca7cdeff07535124d85479085c3803df9 Mon Sep 17 00:00:00 2001
From: Knut Dundas Moraa <knutdundasmoraa@gmail.com>
Date: Tue, 25 Jul 2023 15:21:06 -0400
Subject: [PATCH 15/19] rd drudgery

---
 tests/test_gaussian_model.py | 13 ++++---------
 1 file changed, 4 insertions(+), 9 deletions(-)

diff --git a/tests/test_gaussian_model.py b/tests/test_gaussian_model.py
index f2412a25..afe29001 100644
--- a/tests/test_gaussian_model.py
+++ b/tests/test_gaussian_model.py
@@ -16,9 +16,7 @@ class TestGaussianModel(TestCase):
 
     @classmethod
     def setUp(cls):
-        """
-        Initialise the GaussianModel
-        """
+        # Initialise the GaussianModel
         parameter_definition = {
             'mu': {
                 'fit_guess': 0.,
@@ -39,9 +37,6 @@ def setUp(cls):
             parameter_definition=parameter_definition)
 
     def test_data_generation(self):
-        """
-        Test of generate_data and fit method of the GaussianModel class
-        """
         # test data generation:
         self.simple_model.data = self.simple_model.generate_data(mu=0, sigma=2)
 
@@ -51,7 +46,7 @@ def test_data_storage(self):
         self.simple_model.data = self.simple_model.generate_data(mu=0, sigma=2)
         self.simple_model.store_data(toydata_file, [self.simple_model.data])
         stored_data = inference_interface.toydata_from_file('simple_data.hdf5')
-        assert self.simple_model.data == stored_data[0] , "Stored data disagrees with data!"
+        assert self.simple_model.data == stored_data[0], "Stored data disagrees with data!"
 
         remove("simple_data.hdf5")
 
@@ -61,5 +56,5 @@ def test_fit(self):
         hat_meas = self.simple_model.data[0]["hat_mu"]
         best_fit, lf = self.simple_model.fit(sigma=2)
         hat_fit = best_fit["mu"]
-        np.testing.assert_almost_equal(hat_meas, hat_fit), "best-fit does not agree"
-        np.testing.assert_almost_equal(lf, sps.norm(hat_fit, 2).logpdf(hat_meas)) , "likelihood function disagrees"
\ No newline at end of file
+        np.testing.assert_almost_equal(hat_meas, hat_fit)
+        np.testing.assert_almost_equal(lf, sps.norm(hat_fit, 2).logpdf(hat_meas))

From 258a15b7d8ab11fb8fc88422a14c9cd6ea087c0a Mon Sep 17 00:00:00 2001
From: dachengx <dx2227@columbia.edu>
Date: Wed, 26 Jul 2023 03:51:11 +0800
Subject: [PATCH 16/19] Use setUp instead of __init__ of TestCase

---
 tests/test_blueice_extended_model.py | 16 +++++----
 tests/test_gaussian_model.py         | 49 +++++++++++++---------------
 tests/test_parameter.py              | 10 +++---
 3 files changed, 36 insertions(+), 39 deletions(-)

diff --git a/tests/test_blueice_extended_model.py b/tests/test_blueice_extended_model.py
index a0897478..4ff09ed3 100644
--- a/tests/test_blueice_extended_model.py
+++ b/tests/test_blueice_extended_model.py
@@ -8,12 +8,12 @@
 class TestBlueiceExtendedModel(TestCase):
     """Test of the BlueiceExtendedModel class"""
 
-    def __init__(self, *args, **kwargs):
-        """Initialize the BlueiceExtendedModel class"""
-        super().__init__(*args, **kwargs)
-        self.config = load_yaml('unbinned_wimp_statistical_model.yaml')
-        self.n_likelihood_terms = len(self.config['likelihood_config']['likelihood_terms'])
-        self.set_new_model()
+    @classmethod
+    def setUp(cls):
+        """Initialise the BlueiceExtendedModel instance"""
+        cls.config = load_yaml('unbinned_wimp_statistical_model.yaml')
+        cls.n_likelihood_terms = len(cls.config['likelihood_config']['likelihood_terms'])
+        cls.set_new_model(cls)
 
     def set_new_model(self):
         """Set a new BlueiceExtendedModel instance"""
@@ -57,8 +57,10 @@ def test_fit(self):
         """Test of the fit method"""
         self.model.data = self.model.generate_data()
         fit_result, max_llh = self.model.fit()
-        # TODO: check whether all parameters are in fit_result
+        # TODO:
+        # check whether all parameters are in fit_result
         # and whether fittable parameters are fitted
+        # and whether the results are in boundaries
 
 
 class TestCustomAncillaryLikelihood(TestCase):
diff --git a/tests/test_gaussian_model.py b/tests/test_gaussian_model.py
index afe29001..bc636233 100644
--- a/tests/test_gaussian_model.py
+++ b/tests/test_gaussian_model.py
@@ -1,22 +1,18 @@
-from unittest import TestCase
-
 from os import remove
-
-import inference_interface
-
-import numpy as np
+from unittest import TestCase
 
 import scipy.stats as sps
+import inference_interface
 
 from alea.models import GaussianModel
 
 
 class TestGaussianModel(TestCase):
-    """Test of the Parameters class"""
+    """Test of the GaussianModel class"""
 
     @classmethod
     def setUp(cls):
-        # Initialise the GaussianModel
+        """Initialise the GaussianModel instance"""
         parameter_definition = {
             'mu': {
                 'fit_guess': 0.,
@@ -33,28 +29,27 @@ def setUp(cls):
                 'nominal_value': 1.,
             },
         }
-        cls.simple_model = GaussianModel(
+        cls.model = GaussianModel(
             parameter_definition=parameter_definition)
 
     def test_data_generation(self):
-        # test data generation:
-        self.simple_model.data = self.simple_model.generate_data(mu=0, sigma=2)
+        """Test generation of data"""
+        self.model.data = self.model.generate_data(mu=0, sigma=2)
 
     def test_data_storage(self):
-        # test data store+load:
+        """Test storage of data to file and retrieval of data from file"""
         toydata_file = 'simple_data.hdf5'
-        self.simple_model.data = self.simple_model.generate_data(mu=0, sigma=2)
-        self.simple_model.store_data(toydata_file, [self.simple_model.data])
-        stored_data = inference_interface.toydata_from_file('simple_data.hdf5')
-        assert self.simple_model.data == stored_data[0], "Stored data disagrees with data!"
-
-        remove("simple_data.hdf5")
-
-    def test_fit(self):
-        # test fitting:
-        self.simple_model.data = self.simple_model.generate_data(mu=0, sigma=2)
-        hat_meas = self.simple_model.data[0]["hat_mu"]
-        best_fit, lf = self.simple_model.fit(sigma=2)
-        hat_fit = best_fit["mu"]
-        np.testing.assert_almost_equal(hat_meas, hat_fit)
-        np.testing.assert_almost_equal(lf, sps.norm(hat_fit, 2).logpdf(hat_meas))
+        self.model.data = self.model.generate_data(mu=0, sigma=2)
+        self.model.store_data(toydata_file, [self.model.data])
+        stored_data = inference_interface.toydata_from_file(toydata_file)
+        assert self.model.data == stored_data[0], 'Stored data disagrees with data!'
+        remove(toydata_file)
+
+    def test_fit_result(self):
+        """Test fitting of data"""
+        self.model.data = self.model.generate_data(mu=0, sigma=2)
+        hat_meas = self.model.data[0]['hat_mu'].item()
+        best_fit, lf = self.model.fit(sigma=2)
+        hat_fit = best_fit['mu']
+        self.assertAlmostEqual(hat_meas, hat_fit)
+        self.assertAlmostEqual(lf, sps.norm(hat_fit, 2).logpdf(hat_meas))
diff --git a/tests/test_parameter.py b/tests/test_parameter.py
index 876788f4..de2ff26c 100644
--- a/tests/test_parameter.py
+++ b/tests/test_parameter.py
@@ -8,11 +8,11 @@
 class TestParameters(TestCase):
     """Test of the Parameters class"""
 
-    def __init__(self, *args, **kwargs):
-        """Initialize the BlueiceExtendedModel class"""
-        super().__init__(*args, **kwargs)
-        self.config = load_yaml('unbinned_wimp_statistical_model.yaml')
-        self.parameters = Parameters.from_config(self.config['parameter_definition'])
+    @classmethod
+    def setUp(cls):
+        """Initialise the Parameters instance"""
+        config = load_yaml('unbinned_wimp_statistical_model.yaml')
+        cls.parameters = Parameters.from_config(config['parameter_definition'])
 
     def test_deep_copyable(self):
         """Test of whether Parameters instance can be deepcopied"""

From 7da8d53e6933e834cae75dbf6059c1cbc45c9b9a Mon Sep 17 00:00:00 2001
From: dachengx <dx2227@columbia.edu>
Date: Thu, 27 Jul 2023 00:02:38 +0800
Subject: [PATCH 17/19] Remove url_base, add get_template_folder_list function

---
 alea/__init__.py                      |  7 +------
 alea/models/__init__.py               |  2 --
 alea/models/blueice_extended_model.py | 12 ++----------
 alea/utils.py                         | 25 +++++++++++++++++++++----
 4 files changed, 24 insertions(+), 22 deletions(-)

diff --git a/alea/__init__.py b/alea/__init__.py
index f008bb17..db25e2bf 100644
--- a/alea/__init__.py
+++ b/alea/__init__.py
@@ -1,18 +1,13 @@
 __version__ = '0.0.0'
 
-from . import model
 from .model import *
 
-from . import models
+from .models import *
 
-from . import utils
 from .utils import *
 
-from . import parameters
 from .parameters import *
 
-from . import simulators
 from .simulators import *
 
-from . import template_source
 from .template_source import *
diff --git a/alea/models/__init__.py b/alea/models/__init__.py
index 530eeafa..17b4e8a1 100644
--- a/alea/models/__init__.py
+++ b/alea/models/__init__.py
@@ -1,5 +1,3 @@
-from . import gaussian_model
 from .gaussian_model import *
 
-from . import blueice_extended_model
 from .blueice_extended_model import *
diff --git a/alea/models/blueice_extended_model.py b/alea/models/blueice_extended_model.py
index ddfe47f4..b63070c7 100644
--- a/alea/models/blueice_extended_model.py
+++ b/alea/models/blueice_extended_model.py
@@ -11,7 +11,7 @@
 from alea.model import StatisticalModel
 from alea.parameters import Parameters
 from alea.simulators import BlueiceDataGenerator
-from alea.utils import adapt_likelihood_config_for_blueice
+from alea.utils import adapt_likelihood_config_for_blueice, get_template_folder_list
 
 
 class BlueiceExtendedModel(StatisticalModel):
@@ -107,15 +107,7 @@ def _build_ll_from_config(self, likelihood_config: dict) -> "LogLikelihoodSum":
         """
         lls = []
 
-        if "template_folder" not in likelihood_config:
-            likelihood_config["template_folder"] = []
-        if isinstance(likelihood_config["template_folder"], str):
-            template_folder_list = [likelihood_config["template_folder"]]
-        elif isinstance(likelihood_config["template_folder"], list):
-            template_folder_list = likelihood_config["template_folder"]
-        else:
-            raise ValueError(
-                "template_folder must be either a string or a list of strings.")
+        template_folder_list = get_template_folder_list(likelihood_config)
 
         # Iterate through each likelihood term in the configuration
         for config in likelihood_config["likelihood_terms"]:
diff --git a/alea/utils.py b/alea/utils.py
index 631c87f0..0dc2d30a 100644
--- a/alea/utils.py
+++ b/alea/utils.py
@@ -3,10 +3,12 @@
 import pkg_resources
 from copy import deepcopy
 from pydoc import locate
-from warnings import warn
+import logging
 
 import numpy as np
 
+logging.basicConfig(level=logging.INFO)
+
 
 def get_analysis_space(analysis_space: dict) -> list:
     eval_analysis_space = []
@@ -81,7 +83,7 @@ def get_file_path(fname, folder_list=None):
     Try 5 methods in the following order
 
     #. fname begin with '/', return absolute path
-    #. url_base begin with '/', return url_base + name
+    #. folder begin with '/', return folder + name
     #. can get file from _get_abspath, return alea internal file path
     #. can be found in local installed ntauxfiles, return ntauxfiles absolute path
     #. can be downloaded from MongoDB, download and return cached path
@@ -94,12 +96,12 @@ def get_file_path(fname, folder_list=None):
         return fname
 
     # 2. From local folder
-    # Use url_base as prefix
+    # Use folder as prefix
     for folder in folder_list:
         if folder.startswith('/'):
             fpath = os.path.join(folder, fname)
             if os.path.exists(fpath):
-                warn(f'Load {fname} successfully from {fpath}')
+                logging.info(f'Load {fname} successfully from {fpath}')
                 return fpath
 
     # 3. From alea internal files
@@ -110,3 +112,18 @@ def get_file_path(fname, folder_list=None):
 
     # raise error when can not find corresponding file
     raise RuntimeError(f'Can not find {fname}, please check your file system')
+
+
+def get_template_folder_list(likelihood_config):
+    """Get a list of template_folder from likelihood_config"""
+    if "template_folder" not in likelihood_config:
+        # return empty list if template_folder is not specified
+        likelihood_config["template_folder"] = []
+    if isinstance(likelihood_config["template_folder"], str):
+        template_folder_list = [likelihood_config["template_folder"]]
+    elif isinstance(likelihood_config["template_folder"], list):
+        template_folder_list = likelihood_config["template_folder"]
+    else:
+        raise ValueError(
+            "template_folder must be either a string or a list of strings.")
+    return template_folder_list

From 0250340a58b33d2487c2bd4b8767d01f0bc68fc8 Mon Sep 17 00:00:00 2001
From: dachengx <dx2227@columbia.edu>
Date: Thu, 27 Jul 2023 00:25:17 +0800
Subject: [PATCH 18/19] Add TODO comment on get_expectation_values

---
 alea/models/blueice_extended_model.py | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/alea/models/blueice_extended_model.py b/alea/models/blueice_extended_model.py
index b63070c7..c3afd803 100644
--- a/alea/models/blueice_extended_model.py
+++ b/alea/models/blueice_extended_model.py
@@ -80,6 +80,9 @@ def get_expectation_values(self, **kwargs) -> dict:
         """
         Return total expectation values (summed over all likelihood terms with the same name)
         given a number of named parameters (kwargs)
+        TODO: current implementation is not elegant
+        because it calls the ll and requires data to be set. 
+        We should update this function in the future after we stop using blueice.
         """
         ret = dict()
 

From 34d1c4484eabcb00640b2d338107d21ffe4cb590 Mon Sep 17 00:00:00 2001
From: dachengx <dx2227@columbia.edu>
Date: Thu, 27 Jul 2023 00:28:40 +0800
Subject: [PATCH 19/19] Minor change

---
 alea/models/blueice_extended_model.py | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/alea/models/blueice_extended_model.py b/alea/models/blueice_extended_model.py
index c3afd803..0951f083 100644
--- a/alea/models/blueice_extended_model.py
+++ b/alea/models/blueice_extended_model.py
@@ -80,8 +80,8 @@ def get_expectation_values(self, **kwargs) -> dict:
         """
         Return total expectation values (summed over all likelihood terms with the same name)
         given a number of named parameters (kwargs)
-        TODO: current implementation is not elegant
-        because it calls the ll and requires data to be set. 
+        TODO: Current implementation is not elegant
+        because it calls the ll and requires data to be set.
         We should update this function in the future after we stop using blueice.
         """
         ret = dict()