Add column2Vec test workflow

.github/workflows/py_test.yml

@@ -49,7 +49,7 @@
         run: |
           black --check $(git ls-files '*.py')
 
-  python-tests:
+  similarity-tests:
     env:
       TEST_FILES: tests/similarity_framework/test_similarity*  tests/column2vec/test_column2vec_cache.py tests/runner/test_runner*
     name: Run Python Tests
@@ -91,6 +91,52 @@
       - uses: actions/upload-artifact@v4
         if: github.event_name == 'pull_request'
         with:
-          name: coverage
+          name: Similarity coverage
+          path: coverage.xml
+          retention-days: 1
+
+  column2vec-tests:
+    env:
+      TEST_FILES: tests/column2vec/test_column2vec_cache.py tests/column2vec/test_column2vec.py
+    name: Run Column2Vec Tests
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout repository
+        uses: actions/checkout@v4
+
+      - name: Set up Python
+        uses: actions/setup-python@v5
+        with:
+          python-version: '3.11'
+          cache: 'pip'
+
+      - name: Install dependencies
+        run: |
+          pip install -r requirements.txt
+          pip install coverage pytest pytest-xdist
+
+      - name: Run tests
+        run: coverage run --source='column2Vec' -m pytest -v $TEST_FILES
+
+      - name: Show coverage
+        run: coverage report -m --omit=".*.ipynb"
+
+      - name: Create coverage file
+        if: github.event_name == 'pull_request'
+        run: coverage xml
+
+      - name: Get Cover
+        if: github.event_name == 'pull_request'
+        uses: orgoro/[email protected]
+        with:
+          coverageFile: coverage.xml
+          token: ${{ secrets.GITHUB_TOKEN }}
+          thresholdAll: 0.7
+          thresholdNew: 0.9
+
+      - uses: actions/upload-artifact@v4
+        if: github.event_name == 'pull_request'
+        with:
+          name: Column2Vec coverage
           path: coverage.xml
           retention-days: 1

.gitignore

@@ -122,6 +122,7 @@ celerybeat.pid
 *.sage.py
 
 # Environments
+.env
 .config
 .venv
 env/
@@ -165,6 +166,3 @@ cython_debug/
 # Custom for this project
 fingerprints/
 **/.DS_Store
-
-column2vec/research
-measurement

column2vec/src/functions.py

@@ -9,6 +9,7 @@
 import numpy as np
 import pandas as pd
 import plotly.express as px
+from memory_profiler import profile
 from sentence_transformers import (
     SentenceTransformer,
 )
@@ -41,7 +42,6 @@ def get_nonnumerical_data(
         metadata1 = metadata_creator.get_metadata(MetadataCreatorInput(dataframe=data))
         column_names = metadata1.get_column_names_by_type(NONNUMERICAL)
         for name in column_names:
-            print(f" {i} : {name}")
             result[name + str(index)] = data[name]
     return result
 
@@ -147,9 +147,5 @@ def compute_distances(vectors: dict):
     :param vectors: Dictionary of embedding vectors
     :return: matrix with distances
     """
-    res = {}
-    for key1, vec1 in vectors.items():
-        res[key1] = {}
-        for key2, vec2 in vectors.items():
-            res[key1][key2] = 1 - cosine_sim(vec1, vec2)
+    res = {key1: {key2: 1 - cosine_sim(vec1, vec2) for key2, vec2 in vectors.items()} for key1, vec1 in vectors.items()}
     return res

tests/column2vec/test_column2vec.py

@@ -1,8 +1,10 @@
 import os
+import sys
 import unittest
 import time
 
 import pandas as pd
+from memory_profiler import profile
 from sentence_transformers import SentenceTransformer
 
 from column2vec.src.column2vec import (column2vec_as_sentence, column2vec_as_sentence_clean,
@@ -19,7 +21,16 @@
 # MODEL = 'all-mpnet-base-v2'  # bert-base-nli-mean-tokens
 MODEL = 'sentence-transformers/all-mpnet-base-v2'  #
 THIS_DIR = os.path.dirname(os.path.abspath(__file__))
+TRANSFORMER = SentenceTransformer(MODEL)
 
+def get_size(foo):
+    num =  sys.getsizeof(foo)
+    suffix="B"
+    for unit in ("", "Ki", "Mi", "Gi", "Ti", "Pi", "Ei", "Zi"):
+        if abs(num) < 1024.0:
+            return f"{num:3.1f}{unit}{suffix}"
+        num /= 1024.0
+    return f"{num:.1f}Yi{suffix}"
 
 def vectors_are_same(vec1, vec2):
     for i, j in zip(vec1, vec2):
@@ -61,7 +72,7 @@ def get_data(files):
 class TestSimilarityOfVectors(unittest.TestCase):
     @classmethod
     def setUpClass(cls):
-        cls.model = SentenceTransformer(MODEL)
+        cls.model = TRANSFORMER
         file_m2 = os.path.join(THIS_DIR, os.pardir, 'data/netflix_titles.csv')
         # make an array of all the files
         files = [file_m2]
@@ -77,7 +88,7 @@ def setUpClass(cls):
             stop += 1
 
     def test_column2vec_as_sentence(self):
-        model = SentenceTransformer(MODEL)
+        model = TRANSFORMER
         self.assertTrue(
             vectors_are_same(column2vec_as_sentence(self.first, model, "a"),
                              column2vec_as_sentence(self.first, self.model, "b")))
@@ -88,7 +99,7 @@ def test_column2vec_as_sentence(self):
                              column2vec_as_sentence(self.third, self.model, "f")))
 
     def test_column2vec_as_sentence_clean(self):
-        model = SentenceTransformer(MODEL)
+        model = TRANSFORMER
         self.assertTrue(vectors_are_same(column2vec_as_sentence_clean(self.first, model, "g"),
                                          column2vec_as_sentence_clean(self.first, self.model, "h")))
         self.assertTrue(vectors_are_same(column2vec_as_sentence_clean(self.second, model, "i"),
@@ -97,7 +108,7 @@ def test_column2vec_as_sentence_clean(self):
                                          column2vec_as_sentence_clean(self.third, self.model, "l")))
 
     def test_column2vec_as_sentence_clean_uniq(self):
-        model = SentenceTransformer(MODEL)
+        model = TRANSFORMER
         self.assertTrue(vectors_are_same(column2vec_as_sentence_clean_uniq(self.first, model, "m"),
                                          column2vec_as_sentence_clean_uniq(self.first, self.model, "n")))
         self.assertTrue(vectors_are_same(column2vec_as_sentence_clean_uniq(self.second, model, "o"),
@@ -106,14 +117,14 @@ def test_column2vec_as_sentence_clean_uniq(self):
                                          column2vec_as_sentence_clean_uniq(self.third, self.model, "r")))
 
     def test_column2vec_avg(self):
-        model = SentenceTransformer(MODEL)
+        model = TRANSFORMER
         self.assertTrue(vectors_are_same(column2vec_avg(self.first, model, "v"),
                                          column2vec_avg(self.first, self.model, "s")))
         # self.assertTrue(vectors_are_same(column2vec_avg(self.second, model), column2vec_avg(self.second, self.model)))
         # self.assertTrue(vectors_are_same(column2vec_avg(self.third, model), column2vec_avg(self.third, self.model)))
 
     def test_column2vec_weighted_avg(self):
-        model = SentenceTransformer(MODEL)
+        model = TRANSFORMER
         self.assertTrue(vectors_are_same(column2vec_weighted_avg(self.first, model, "u"),
                                          column2vec_weighted_avg(self.first, self.model, "w")))
         # self.assertTrue(vectors_are_same(column2vec_weighted_avg(self.second, model),
@@ -122,12 +133,12 @@ def test_column2vec_weighted_avg(self):
         # column2vec_weighted_avg(self.third, self.model)))
 
     def test_column2vec_sum(self):
-        model = SentenceTransformer(MODEL)
+        model = TRANSFORMER
         self.assertTrue(vectors_are_same(column2vec_sum(self.first, model, "x"),
                                          column2vec_sum(self.first, self.model, "y")))
 
     def test_column2vec_weighted_sum(self):
-        model = SentenceTransformer(MODEL)
+        model = TRANSFORMER
         self.assertTrue(vectors_are_same(column2vec_weighted_sum(self.first, model, "z"),
                                          column2vec_weighted_sum(self.first, self.model, "ab")))
 
@@ -269,6 +280,9 @@ def test_column2vec_as_sentence_clean_uniq(self):
 
     @unittest.skipIf(SKIP_SIMILAR, "Skipping test_column2vec_avg ...")
     def test_column2vec_avg(self):
+        # slow one 1:30
+        # slow one 1:30
+        # slow one 1:30
         vectors_sentence = get_vectors(column2vec_avg, self.data)
         distances = compute_distances(vectors_sentence)
         self.print_accuracy(distances, "AVG")

tests/column2vec/test_column2vec_cache.py

@@ -31,8 +31,7 @@ def setUpClass(cls):
         for i in cls.data:
             cls.first = cls.data[i].head(100)
             break
-        cache.set_file("generated/test.csv")
-        cls.model = SentenceTransformer(MODEL)
+        cache.set_file("cache_test2.csv")
 
     def setUp(self):
         cache.clear_cache()
@@ -47,7 +46,6 @@ def test_column2vec_as_sentence(self):
         cache.off()
         third = time_measure_function(column2vec_as_sentence, self.first, self.model,  "a")
 
-        print(f"{first} : {second} : {third}")
         self.assertGreater(first, second)
         self.assertGreater(third, second)
 
@@ -60,7 +58,6 @@ def test_column2vec_as_sentence_clean(self):
         cache.off()
         third = time_measure_function(column2vec_as_sentence_clean, self.first, self.model, "a")
 
-        print(f"{first} : {second} : {third}")
         self.assertGreater(first, second)
         self.assertGreater(third, second)
 
@@ -72,7 +69,6 @@ def test_column2vec_as_sentence_clean_uniq(self):
         cache.off()
         third = time_measure_function(column2vec_as_sentence_clean_uniq, self.first, self.model, "a")
 
-        print(f"{first} : {second} : {third}")
         self.assertGreater(first, second)
         self.assertGreater(third, second)
 
@@ -84,7 +80,6 @@ def test_column2vec_avg(self):
         cache.off()
         third = time_measure_function(column2vec_avg, self.first, self.model, "a")
 
-        print(f"{first} : {second} : {third}")
         self.assertGreater(first, second)
         self.assertGreater(third, second)
 
@@ -96,8 +91,6 @@ def test_column2vec_weighted_avg(self):
         cache.off()
         third = time_measure_function(column2vec_weighted_avg, self.first, self.model, "a")
 
-        print(f"{first} : {second} : {third}")
-
         self.assertGreater(first, second)
         self.assertGreater(third, second)
 
@@ -109,8 +102,6 @@ def test_column2vec_sum(self):
         cache.off()
         third = time_measure_function(column2vec_sum, self.first, self.model, "a")
 
-        print(f"{first} : {second} : {third}")
-
         self.assertGreater(first, second)
         self.assertGreater(third, second)
 
@@ -123,8 +114,6 @@ def test_column2vec_weighted_sum(self):
         cache.off()
         third = time_measure_function(column2vec_weighted_sum, self.first, self.model, "a")
 
-        print(f"{first} : {second} : {third}")
-
         self.assertGreater(first, second)
         self.assertGreater(third, second)
 
@@ -140,13 +129,12 @@ def setUpClass(cls):
         cls.data = get_nonnumerical_data(files)
         skip = True
         for i in cls.data:
-            if skip:
+            if skip:  # skip first column to get different than in previous test
                 skip = False
                 continue
             cls.first = cls.data[i].head(100)
             break
         cache.set_file("cache_test.csv")
-        cls.model = SentenceTransformer(MODEL)
 
     def setUp(self):
         cache.clear_cache()
@@ -163,7 +151,6 @@ def test_column2vec_as_sentence(self):
         cache.off()
         third = time_measure_function(column2vec_as_sentence, self.first, self.model,  "a")
 
-        print(f"{first} : {second} : {third}")
         self.assertGreater(first, second)
         self.assertGreater(third, second)
 
@@ -178,7 +165,6 @@ def test_column2vec_as_sentence_clean(self):
         cache.off()
         third = time_measure_function(column2vec_as_sentence_clean, self.first, self.model, "a")
 
-        print(f"{first} : {second} : {third}")
         self.assertGreater(first, second)
         self.assertGreater(third, second)
 
@@ -192,7 +178,6 @@ def test_column2vec_as_sentence_clean_uniq(self):
         cache.off()
         third = time_measure_function(column2vec_as_sentence_clean_uniq, self.first, self.model, "a")
 
-        print(f"{first} : {second} : {third}")
         self.assertGreater(first, second)
         self.assertGreater(third, second)
 
@@ -206,7 +191,6 @@ def test_column2vec_avg(self):
         cache.off()
         third = time_measure_function(column2vec_avg, self.first, self.model, "a")
 
-        print(f"{first} : {second} : {third}")
         self.assertGreater(first, second)
         self.assertGreater(third, second)
 
@@ -220,7 +204,6 @@ def test_column2vec_weighted_avg(self):
         cache.off()
         third = time_measure_function(column2vec_weighted_avg, self.first, self.model, "a")
 
-        print(f"{first} : {second} : {third}")
 
         self.assertGreater(first, second)
         self.assertGreater(third, second)
@@ -230,13 +213,9 @@ def test_column2vec_sum(self):
         first = time_measure_function(column2vec_sum, self.first, self.model, "a")
         cache.save_persistently()
         cache.clear_cache()
-
         second = time_measure_function(column2vec_sum, self.first, self.model, "a")
         cache.off()
         third = time_measure_function(column2vec_sum, self.first, self.model, "a")
-
-        print(f"{first} : {second} : {third}")
-
         self.assertGreater(first, second)
         self.assertGreater(third, second)
 
@@ -251,8 +230,6 @@ def test_column2vec_weighted_sum(self):
         cache.off()
         third = time_measure_function(column2vec_weighted_sum, self.first, self.model, "a")
 
-        print(f"{first} : {second} : {third}")
-
         self.assertGreater(first, second)
         self.assertGreater(third, second)