eyaltrabelsi

def test_get_family_size():
   df = get_production_sample(pct=0.1)
   func = lambda x: get_family_size(x['SibSp'], x['Parch'])
   sizes = df.apply(func, axis=1)

   assert all(size >= 1 for size in sizes)

eyaltrabelsi

@given(st.integers(), st.integers())
@example(-1, 0)
def test_get_family_size(sib, parch):
   fs = get_family_size(sib, parch)
   assert fs >= 1

eyaltrabelsi

@given(st.integers(), st.integers())
def test_get_family_size(sib, parch):
  fs = get_family_size(sib, parch)
  assert fs >= 1

eyaltrabelsi

def get_family_size(sib, parch):
   return sib_sp + parch + 1

eyaltrabelsi

def test_get_family_size():
   sib = 2
   parent = 2
   actual = get_family_size(sib, parent)
   assert actual == 5

eyaltrabelsi

def is_royal(name):
   title = name.split(",")[1].split(".")[0]
   title = parsed_name.strip()
    return title in ['the Countess', 'Lady', 'Sir']

eyaltrabelsi

def test_queries_synatx():
  queries = glob.glob('**/*.sql',recursive=True)
  engine = create_engine('<COOL_CONNECTION_STR>')
  with engine.connect() as con:
  for query_file in queries_list:
  with open(query_file) as f:
  query = f.read()
  test_query = f"Explain {query}"
  con.execute(test_query)

eyaltrabelsi

def test_model_performance():
  model = get_trained_model(model)
  base = get_production_sample(pct=0.1)
  base = feature_engineering(input)
  base_preds = model.predict(base)
  labels = get_true_labels(base)
  auc = roc_auc_score(true_labels, base_preds)
  assert auc > 0.8

eyaltrabelsi

def test_inference_latency():
  model = get_trained_model(model)
  base = get_production_sample(pct=0.1)
  base = feature_engineering(input)
  start_time = time.time()
  base_preds = model.predict(base)
  end_time = time.time()
  assert (end_time - start_time) <= 200

eyaltrabelsi

def test_model_bias():
  model = get_trained_model(model)
  base = get_production_sample(pct=0.1)
  base = feature_engineering(input)
  base_preds = model.predict(base)

  embark_b = base.copy()
  embark_b["Embarked"] = "B"
  embark_b_preds = model.predict(embark_b)
  assert average(abs(embark_b_preds - base_preds)) < 0.1