tests: add strategies and data dictionary tests

tests/conftest.py

@@ -0,0 +1,20 @@
+"Common fixtures"
+
+from pathlib import Path
+
+import pytest
+
+import arcmapper
+
+dictionary_file = str(Path(__file__).parent / "data" / "ccpuk_dictionary.csv")
+arc_file = str(Path(__file__).parent / "data" / "ARCH.csv")
+
+@pytest.fixture(scope="session")
+def arc_schema():
+    return arcmapper.read_arc_schema(arc_file)
+
+@pytest.fixture(scope="session")
+def data_dictionary():
+    return arcmapper.read_data_dictionary(dictionary_file, description_field="Field Label",
+                              response_field="Choices, Calculations, OR Slider Labels", response_func="redcap")
+

tests/test_dictionary.py

@@ -0,0 +1,39 @@
+import pandas as pd
+
+from arcmapper.dictionary import read_from_jsonschema
+
+
+EXAMPLE_JSON_SCHEMA = """{
+  "$schema": "http://json-schema.org/draft-07/schema#",
+  "type": "object",
+  "properties": {
+    "category": {
+      "type": "string",
+      "enum": ["option1", "option2", "option3"],
+      "description": "category of the variable"
+    },
+    "value": {
+      "type": "number",
+      "description": "default value of variable"
+    },
+    "full_name": {
+      "type": "string",
+      "description": "full name of variable"
+    }
+  },
+  "required": ["category", "value", "full_name"]
+}"""
+
+def test_read_data_dictionary(data_dictionary):
+    assert data_dictionary.columns.tolist()== ["variable", "description", "responses", "type"]
+
+
+def test_read_from_jsonschema():
+    dd = read_from_jsonschema(EXAMPLE_JSON_SCHEMA)
+    print(dd)
+    expected = pd.DataFrame({'variable': ["category", "value", "full_name"],
+                             'description': ['category of the variable', 'default value of variable', 'full name of variable'],
+                             'responses': [[('option1', 'option1'),('option2', 'option2'),('option3', 'option3')], None, None],
+                             'type': ['categorical', 'number', 'string']})
+    assert dd.equals(expected)
+

tests/test_strategies.py

@@ -0,0 +1,19 @@
+from pathlib import Path
+import numpy as np
+
+from arcmapper.strategies import map, get_categorical_mapping
+
+dictionary_file = Path(__file__).parent / "data" / "ccpuk_dictionary.csv"
+arc_file = Path(__file__).parent / "data" / "ARCH.csv"
+
+def test_get_categorical_mapping():
+    sim = np.array([[0.1, 0.8], [0.9, 0.1]])
+    target = ["male", "female"]
+    source = ["femme", "homme"]
+    assert get_categorical_mapping(source, target, sim) == {"femme": "female", "homme": "male"}
+
+
+def test_tf_idf(data_dictionary, arc_schema):
+    map("tf-idf", data_dictionary, arc_schema, num_matches=3)
+
+