Add Components to Create Features from URL and Email Logical Types #2550

freddyaboulton · 2021-07-23T20:26:17Z

Pull Request Description

build_conda_pkg will fail until after we merge cause we need to bump the lower bound on feature tools to 0.26.1

codecov · 2021-07-23T20:29:21Z

Codecov Report

Merging #2550 (62a4313) into main (1f0b96d) will increase coverage by 0.1%.
The diff coverage is 99.6%.

@@           Coverage Diff           @@
##            main   #2550     +/-   ##
=======================================
+ Coverage   99.9%   99.9%   +0.1%     
=======================================
  Files        285     287      +2     
  Lines      26115   26338    +223     
=======================================
+ Hits       26081   26304    +223     
  Misses        34      34

Impacted Files	Coverage Δ
evalml/pipelines/components/__init__.py	`100.0% <ø> (ø)`
...alml/pipelines/components/transformers/__init__.py	`100.0% <ø> (ø)`
evalml/tests/component_tests/test_utils.py	`97.5% <66.7%> (ø)`
.../components/transformers/preprocessing/__init__.py	`100.0% <100.0%> (ø)`
...rs/preprocessing/transform_primitive_components.py	`100.0% <100.0%> (ø)`
evalml/pipelines/utils.py	`99.1% <100.0%> (+0.1%)`	⬆️
...nt_tests/test_ft_transform_primitive_components.py	`100.0% <100.0%> (ø)`
evalml/tests/conftest.py	`98.6% <100.0%> (+0.1%)`	⬆️
...s/prediction_explanations_tests/test_explainers.py	`100.0% <100.0%> (ø)`
...understanding_tests/test_permutation_importance.py	`100.0% <100.0%> (ø)`
... and 3 more

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 1f0b96d...62a4313. Read the comment docs.

freddyaboulton · 2021-07-26T15:40:10Z

evalml/pipelines/components/transformers/preprocessing/transform_primitive_components.py

+from evalml.utils import infer_feature_types
+
+
+class _ExtractFeaturesWithTransformPrimitives(Transformer):


Added a parent class to minimize the code duplication that comes from the following pattern:

specify some ft primitives at class definition

Run dfs to extract features with those primitives

Delete original columns

Maybe this could be used for the TextFeaturizer later as well. But if we find this abstraction useful, we may need to rethink how the casting to categorical at the end of transform happens.

I think this is a great idea

❤️ this!

ParthivNaresh

Awesome work, looks great to me. Thanks for setting up a parent class for feature extraction, I have no doubt it'll come in handy. Love the tests too!

ParthivNaresh · 2021-07-27T14:59:48Z

evalml/pipelines/components/transformers/preprocessing/transform_primitive_components.py

+    """{}"""
+
+    def __init__(self, random_seed=0, **kwargs):
+        self._primitives_provenance = {}


Does this variable get used?

ParthivNaresh · 2021-07-27T15:07:55Z

evalml/pipelines/components/transformers/preprocessing/transform_primitive_components.py

+from evalml.utils import infer_feature_types
+
+
+class _ExtractFeaturesWithTransformPrimitives(Transformer):


I think this is a great idea

ParthivNaresh · 2021-07-27T15:15:09Z

evalml/tests/component_tests/test_utils.py

    elif is_running_py_39_or_above:
-        n_components = 47
-    else:
        n_components = 49


Kind of confused as to why this isn't throwing a codecov warning for lack of coverage. Did Codecov not check this change because n_components = 49 seems to have come from the else statement before? That's wild

ParthivNaresh · 2021-07-27T15:28:57Z

evalml/tests/component_tests/test_ft_transform_primitive_components.py

+    ],
+)
+def test_component_fit_transform(
+    component, make_data, make_expected, make_expected_ltypes, df_with_url_and_email


Love the way you parametrized this with helper functions

ParthivNaresh · 2021-07-27T15:31:20Z

evalml/tests/model_understanding_tests/prediction_explanations_tests/test_explainers.py

+    )
+    assert (
+        not pd.Series(
+            explanations["explanations"][1]["explanations"][0][


After reading this test, the word explanations is starting to sound weird in my head

eccabay

Nice, this is all super clean! I think the parent class was a great idea

eccabay · 2021-07-27T15:14:45Z

evalml/pipelines/components/transformers/preprocessing/transform_primitive_components.py

+        # featuretools expects str-type column names
+        X_to_transform.rename(columns=str, inplace=True)
+        ft_variable_types = self._get_feature_types_for_featuretools(X)
+        # all_email_variable_types =


Leftover comment?

Yes! Will delete thank you

eccabay · 2021-07-27T15:39:49Z

evalml/tests/component_tests/test_ft_transform_primitive_components.py

+def test_url_featurizer_init():
+    url = URLFeaturizer()
+    assert url.parameters == {}
+
+
+def test_email_featurizer_init():
+    email = EmailFeaturizer()
+    assert email.parameters == {}


nit: it might be cleaner/more scalable to parameterize these

chukarsten

This is extremely thorough. I'm very glad for your complete product knowledge as a lot of these extra tests I would not have known to add. Looks good.

chukarsten · 2021-07-27T15:10:09Z

docs/source/release_notes.rst

 **Future Release**
    * Enhancements
+        * Added components to extract features from ``URL`` and ``EmailAddress`` Logical Types :pr:`2550`
+        * Added support for `NaN` values in ``TextFeaturizer`` :pr:`2532`


chukarsten · 2021-07-27T15:20:01Z

evalml/pipelines/components/transformers/preprocessing/transform_primitive_components.py

+from evalml.utils import infer_feature_types
+
+
+class _ExtractFeaturesWithTransformPrimitives(Transformer):


chukarsten · 2021-07-28T14:04:31Z

evalml/pipelines/components/transformers/preprocessing/transform_primitive_components.py

+        provenance = {}
+        for feature in features:
+            input_col = feature.base_features[0].get_name()
+            # Return a copy because `get_feature_names` returns a reference to the names
+            output_features = [name for name in feature.get_feature_names()]
+            if input_col not in provenance:
+                provenance[input_col] = output_features
+            else:
+                provenance[input_col] += output_features
+        return provenance


Shame this is different from the component graph provenance, but nothing to do there.

chukarsten · 2021-07-28T14:06:24Z

evalml/pipelines/utils.py


+    email_columns = list(X.ww.select("EmailAddress", return_schema=True).columns)
+    if len(email_columns) > 0:
+        pp_components.append(EmailFeaturizer)


chukarsten · 2021-07-28T14:16:20Z

evalml/tests/conftest.py

+                "abalone_0@gmail.com",
+                "AbaloneRings@yahoo.com",
+                "abalone_2@abalone.com",
+                "$titanic_data%&@hotmail.com",


these are the most depressing email addresses I've ever seen. I'm tempted to try to register the hotmail one

freddyaboulton force-pushed the 2444-email-url branch from e14e712 to 143eafa Compare July 23, 2021 20:27

freddyaboulton force-pushed the 2444-email-url branch from 469ee09 to f24e7ba Compare July 26, 2021 14:47

freddyaboulton commented Jul 26, 2021

View reviewed changes

freddyaboulton requested a review from a team July 26, 2021 16:42

freddyaboulton marked this pull request as ready for review July 26, 2021 16:42

auto-assign bot assigned freddyaboulton Jul 26, 2021

freddyaboulton force-pushed the 2444-email-url branch from e92d642 to 7837cd9 Compare July 26, 2021 21:57

ParthivNaresh approved these changes Jul 27, 2021

View reviewed changes

eccabay approved these changes Jul 27, 2021

View reviewed changes

freddyaboulton force-pushed the 2444-email-url branch from a3fe6e9 to 97dbb04 Compare July 27, 2021 18:56

chukarsten approved these changes Jul 28, 2021

View reviewed changes

freddyaboulton added 11 commits July 28, 2021 12:08

Adding tests for email featurizer

fb4f956

Implementation with tests

0a5e9fe

Adding model understanding tests

263b595

Add to rl and rebase

2f1286d

Pin ft min version at 0,26.0

85d49e0

Adding tests for make_pipeline

7a6a374

Linting

758d2f2

Deleting transform method. Docstring not that crucial

d6fda10

Fix 2532 release notes

bfb8187

Fix test broke during conflict

0eba764

Address comments

62a4313

freddyaboulton force-pushed the 2444-email-url branch from 97dbb04 to 62a4313 Compare July 28, 2021 16:14

freddyaboulton merged commit aae7b55 into main Jul 28, 2021

freddyaboulton deleted the 2444-email-url branch July 28, 2021 17:22

chukarsten mentioned this pull request Aug 3, 2021

Release v0.30.0 #2584

Merged

freddyaboulton mentioned this pull request Nov 2, 2021

EPIC Supporting All WW Types #2443

Open

		from evalml.utils import infer_feature_types


		class _ExtractFeaturesWithTransformPrimitives(Transformer):

Add Components to Create Features from URL and Email Logical Types #2550

Add Components to Create Features from URL and Email Logical Types #2550

Uh oh!

Conversation

freddyaboulton commented Jul 23, 2021 • edited by gsheni Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Pull Request Description

Uh oh!

codecov bot commented Jul 23, 2021 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Codecov Report

Uh oh!

freddyaboulton Jul 26, 2021 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

ParthivNaresh left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

eccabay left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

chukarsten left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

6 participants

freddyaboulton commented Jul 23, 2021 •

edited by gsheni

Loading

codecov bot commented Jul 23, 2021 •

edited

Loading

freddyaboulton Jul 26, 2021 •

edited

Loading