test is broken but trying to write fixtures for creating and removing…

… directories

test/test_io.py

@@ -0,0 +1,102 @@
+"""
+Test the io module
+"""
+import pytest
+import unittest
+import numpy as np
+import os
+import shutil
+from scripts.io import ModelLoader, DataLoader, DataPreparation
+
+# first define some fixtures
+# fixtures are useful for cases where you need to reuse something
+# multiple times in testing
+
+
+@pytest.fixture
+def setup_dir():
+    # create a temporary directory for the saved model
+    dir = "data/"
+    os.makedirs(dir)
+    yield dir
+
+#@pytest.fixture
+#def teardown(path="data/"):
+#    shutil.rmtree(path)
+
+
+def test_modelloader():
+    model = ModelLoader()
+    assert 0 == 0
+
+
+def test_dataloader():
+    data = DataLoader()
+    assert 0 != 1
+
+
+def test_datapreparation():
+    data = DataPreparation()
+    size_df = 1000
+    noise = 'vhigh'
+    data.sample_params_from_prior(size_df)
+    if noise == 'low':
+        sigma = 1
+    if noise == 'medium':
+        sigma = 5
+    if noise == 'high':
+        sigma = 10
+    if noise == 'vhigh':
+        sigma = 100
+    data.simulate_data(data.params,
+                       sigma,
+                       'linear_homogeneous'
+                       )
+    print('shape output', np.shape(data.output[0]))
+    data.output
+    print('shape params', np.shape(data.params))
+    assert np.shape(data.input) != np.shape(data.output), \
+        f"shape of input {np.shape(data.input)} does not match \
+              shape of output {np.shape(data.output)}"
+
+
+def test_datapreparation_and_saver(setup_dir):
+    data = DataPreparation()
+    size_df = 1000
+    noise = 'vhigh'
+    data.sample_params_from_prior(size_df)
+    if noise == 'low':
+        sigma = 1
+    if noise == 'medium':
+        sigma = 5
+    if noise == 'high':
+        sigma = 10
+    if noise == 'vhigh':
+        sigma = 100
+    data.simulate_data(data.params,
+                       sigma,
+                       'linear_homogeneous'
+                       )
+    print('shape output', np.shape(data.output[0]))
+    data.output
+    print('shape params', np.shape(data.params))
+    assert np.shape(data.input) != np.shape(data.output), \
+        f"shape of input {np.shape(data.input)} does not match \
+              shape of output {np.shape(data.output)}"
+    datadict = data.get_dict()
+    assert isinstance(datadict, dict)
+    saver = DataLoader()
+    # save the dataframe
+    name = 'linear_sigma_'+str(sigma) + '_size_'+str(size_df)
+    saver.save_data_h5(name, datadict, path="data/")
+    assert os.path.exists(name+".h5"), f"File {name} does not exist"
+    #shutil.rmtree("data/")
+    #teardown(path="data/")
+
+
+
+"""
+To run this suite of tests, run 'pytest' in the main directory
+"""
+if __name__ == "__main__":
+    unittest.main()