Merge pull request #31 from MetaSys-LISBP/dev

Added Continuous integration

.github/workflows/tests.yml

@@ -0,0 +1,27 @@
+name: Launch tests
+
+on:
+  pull_request:
+    types: [opened, synchronize]
+
+jobs:
+  test:
+    runs-on: ${{ matrix.os}}
+    strategy:
+      matrix:
+        os: [ubuntu-latest, windows-latest, macos-latest]
+        python-version: [ '3.8', '3.9', '3.10', '3.11']
+
+    steps:
+      - uses: actions/checkout@v3
+      - name: Set up Python ${{matrix.python-version}}
+        uses: actions/setup-python@v4
+        with:
+          python-version: ${{matrix.python-version}}
+          architecture: x64
+      - name: Install Dependencies
+        run:
+          python -m pip install --upgrade pip
+          pip install tox tox-gh-actions
+      - name: Test with tox
+        run: tox

physiofit/base/io.py

@@ -231,11 +231,12 @@ def initialize_fitter(self, data, **kwargs):
         """
         Initialize a PhysioFitter object
 
+        :param data: input data
         :param kwargs: Keyword arguments for fitter initialization
         :return: None
         """
-
-        wrong_keys = []
+        #
+        # wrong_keys = []
 
         # Initialize fitter
         fitter = PhysioFitter(

physiofit/tests/conftest.py

@@ -8,23 +8,47 @@
 the fixture function's name.
 """
 
+import os
+
+import pandas as pd
 import pytest
+from numpy import nan
 from physiofit.base.io import IoHandler
 from physiofit.models.base_model import StandardDevs
 
 @pytest.fixture
 def data():
-    """Test data to use in tests"""
+    """Test data to use in tests (taken from Bergès et al., 2021 --> KEIO_ROBOT1_1)"""
 
-    return IoHandler.read_data(
-        r"C:\Users\legregam\PycharmProjects\PhysioFit\physiofit\data\data_example.tsv"
+    return pd.DataFrame.from_dict(
+        {
+            "time": [0, 1.18, 2.27, 3.13, 3.77, 4.42, 4.82, 0.67, 1.72, 2.8, 3.63, 4.27, 4.88],
+            "X": [0.03, 0.05, 0.08, 0.13, 0.18, 0.24, 0.34, nan, nan, nan, nan, nan, nan],
+            "Glucose": [nan, nan, nan, nan, nan, nan, nan, 14, 15.30, 13.68, 12.81, 12.15, 10.93],
+            "Acetate": [nan, nan, nan, nan, nan, nan, nan, 0.01, 0.33, 0.72, 1.17, 1.63, 2.14]
+        }
     )
 
+# time	X	Glc	Ace
+# 0	0.031752	NA	NA
+# 1.18888888888889	0.05292	NA	NA
+# 2.27694444444444	0.084672	NA	NA
+# 3.12833333333333	0.134568	NA	NA
+# 3.77138888888889	0.175392	NA	NA
+# 4.41555555555556	0.244944	NA	NA
+# 4.82277777777778	0.337176	NA	NA
+# 0.0666666666666667	NA	14.0042771787617	0.0130592124021829
+# 1.71666666666667	NA	15.3032563161238	0.325073443768164
+# 2.8	NA	13.6751055268759	0.722570137805818
+# 3.63333333333333	NA	12.8110643342243	1.16940283499877
+# 4.26666666666667	NA	12.1520761714542	1.62777664140305
+# 4.88333333333333	NA	10.9297935900362	2.13639255382759
+
 @pytest.fixture
 def sds():
     return StandardDevs(
         X = 0.2,
-        Glc = 0.5,
-        Ace = 0.5
+        Glucose = 0.5,
+        Acetate = 0.5
     )
 

physiofit/tests/test_base.py

@@ -0,0 +1,110 @@
+"""
+Test the creation and use of the PhysioFitter
+"""
+import numpy as np
+import pandas as pd
+import pytest
+import physiofit
+
+
+def test_physiofitter(data, sds):
+    """
+    Test that the model and PhysioFitter can be safely instanciated from IoHandler
+    """
+
+    io = physiofit.base.io.IoHandler()
+    model = io.select_model("Steady-state batch model", data)
+    assert isinstance(model, physiofit.models.base_model.Model)
+    model.get_params()
+    fitter = io.initialize_fitter(
+        model.data,
+        model=model,
+        sd=sds,
+        debug_mode=True
+    )
+    assert isinstance(fitter, physiofit.base.fitter.PhysioFitter)
+
+def test_simulation(data, sds):
+
+    io = physiofit.base.io.IoHandler()
+    model = io.select_model("Steady-state batch model", data)
+    model.get_params()
+    sim_mat = model.simulate(
+        [param for param in model.parameters_to_estimate.values()],
+        model.experimental_matrix,
+        model.time_vector,
+        model.fixed_parameters
+    )
+    assert isinstance(sim_mat, np.ndarray)
+
+def test_wrong_entry_for_model_data(data, sds):
+
+    with pytest.raises(AttributeError):
+        io = physiofit.base.io.IoHandler()
+        model = io.select_model("Steady-state batch model", None)
+    with pytest.raises(AttributeError):
+        model = io.select_model(
+            "Steady-state batch model",
+            np.array(
+                [[0, 1, 2],[0, 1, 2]]
+            )
+        )
+    with pytest.raises(AttributeError):
+        model = io.select_model(
+            "Steady-state batch model",
+            "Hello world this is an error"
+        )
+
+def test_optimization_process(data, sds):
+
+    io = physiofit.base.io.IoHandler()
+    model = io.select_model("Steady-state batch model", data)
+    model.get_params()
+    fitter = io.initialize_fitter(
+        model.data,
+        model=model,
+        sd=sds,
+        debug_mode=True
+    )
+    fitter.optimize()
+    assert isinstance(fitter.simulated_data, pd.DataFrame)
+    for col in ["X", "Glucose", "Acetate"]:
+        assert col in fitter.simulated_data.columns
+
+def test_monte_carlo(data, sds):
+    io = physiofit.base.io.IoHandler()
+    model = io.select_model("Steady-state batch model", data)
+    model.get_params()
+    fitter = io.initialize_fitter(
+        model.data,
+        model=model,
+        sd=sds,
+        debug_mode=True
+    )
+    fitter.optimize()
+    fitter.monte_carlo_analysis()
+    assert hasattr(fitter, "matrices_ci")
+    assert isinstance(fitter.matrices_ci["lower_ci"], np.ndarray)
+    assert isinstance(fitter.matrices_ci["higher_ci"], np.ndarray)
+    assert np.subtract(fitter.matrices_ci["higher_ci"], fitter.matrices_ci["lower_ci"]).all() > 0
+    assert hasattr(fitter, "parameter_stats")
+    assert isinstance(fitter.parameter_stats, dict)
+
+def test_that_simulated_and_experimental_matrices_are_close(data, sds):
+    io = physiofit.base.io.IoHandler()
+    model = io.select_model("Steady-state batch model", data)
+    model.get_params()
+    fitter = io.initialize_fitter(
+        model.data,
+        model=model,
+        sd=sds,
+        debug_mode=True
+    )
+    fitter.optimize()
+    assert np.allclose(
+        a=fitter.experimental_matrix,
+        b=fitter.simulated_data,
+        rtol=1,
+        equal_nan=True
+    )
+