Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Add DenseClus Implementation notebook for jumpstart #60

Merged
merged 17 commits into from
Feb 29, 2024
Merged

Add DenseClus Implementation notebook for jumpstart #60

Show file tree
Hide file tree
Changes from 12 commits
Commits
Show all changes
17 commits
Select commit Hold shift + click to select a range
5557102
add jumpstart notebook
srushtii-aws Nov 24, 2023
9d01eef
updating with latest changes
momonga-ml Dec 9, 2023
d372620
Merge pull request #51 from awslabs/main
srushtii-aws Dec 17, 2023
907823f
add jumpstart notebook
srushtii-aws Dec 22, 2023
15d008c
add updated nb
srushtii-aws Dec 22, 2023
3609819
clean nb
srushtii-aws Dec 22, 2023
1ee8c4e
update based on feedback
srushtii-aws Dec 28, 2023
857a36a
clleanup
srushtii-aws Dec 28, 2023
47e05da
notebook linter formatting updates
bharven Jan 17, 2024
bdf51d0
incorporate feedback
srushtii-aws Jan 19, 2024
d42ddeb
updated results and analysis
srushtii-aws Feb 6, 2024
78c7330
notebook cleanup
bharven Feb 16, 2024
6dec37b
Updating metircs for clustering scoring; bumping versions
momonga-ml Dec 9, 2023
86cde63
added CH score to evalutor, notebook cleanup
bharven Feb 27, 2024
4f1ba55
notebook linter updates
bharven Feb 27, 2024
0a9946c
combining nbs; version bumps
momonga-ml Feb 29, 2024
4d36fe0
Merge branch 'main' into feature/jumpstart_notebook
momonga-ml Feb 29, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
14 changes: 9 additions & 5 deletions denseclus/DenseClus.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -593,15 +593,15 @@ def predict(self, df_new: pd.DataFrame) -> np.ndarray:
)
return predictions

def evaluate(self) -> np.array:
def evaluate(self, log_dbcv=False) -> np.array:
"""Evaluates the cluster and returns the cluster assigned to each row.

This is a wrapper function for HDBSCAN. It outputs the cluster labels
that HDBSCAN converged on.

Parameters
----------
None : None
log_dbcv (bool) : Whether to log DBCV scores. Defaults to False

Returns
-------
Expand All @@ -612,14 +612,18 @@ def evaluate(self) -> np.array:
clustered = labels >= 0

if isinstance(self.hdbscan_, dict) or self.umap_combine_method == "ensemble":
print(f"DBCV score {self.hdbscan_['hdb_numerical'].relative_validity_}")
print(f"DBCV score {self.hdbscan_['hdb_categorical'].relative_validity_}")
if log_dbcv:
print(f"DBCV numerical score {self.hdbscan_['hdb_numerical'].relative_validity_}")
print(
f"DBCV categorical score {self.hdbscan_['hdb_categorical'].relative_validity_}"
)
embedding_len = self.numerical_umap_.embedding_.shape[0]
coverage = np.sum(clustered) / embedding_len
print(f"Coverage {coverage}")
return labels

print(f"DBCV score {self.hdbscan_.relative_validity_}")
if log_dbcv:
print(f"DBCV score {self.hdbscan_.relative_validity_}")
embedding_len = self.mapper_.embedding_.shape[0]
coverage = np.sum(clustered) / embedding_len
print(f"Coverage {coverage}")
Expand Down
2 changes: 2 additions & 0 deletions notebooks/02_TuningWithHDBSCAN.ipynb
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -365,6 +365,8 @@
],
"source": [
"# we will make our own scorer for DBCV\n",
"\n",
"\n",
"def dbcv_score(X, labels):\n",
" return validity_index(X, labels)\n",
"\n",
Expand Down
2,191 changes: 2,191 additions & 0 deletions notebooks/DenseClusImplentation.ipynb
View file
Open in desktop
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

FYI can't leave comments on specific lines because diff is too big (and github vs code extension doesn't support commenting on notebooks microsoft/vscode-pull-request-github#3462). Will try and be descriptive in comments.

Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

add a comment explaining why we only take native_country = " United-States"

Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Add a markdown description about what/why we are doing in "Create UMAP embeddings & Fit HdbScan for Numerical and Categorical features separately" section (ie "a seemingly straightforward approach may be to try clustering numerical and categorical features separately. lets use this as a baseline to compare against.. ")

also include brief overview of what UMAP and HDBSCAN are - can probably pull this from other notebooks

Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

thoughts about baseline separate numerical/categorical cluster analysis:

  • cluster results don't look super meaningful, could this be improved with hyperparameter optimization? this may be too much for this notebook though, especially considering this is just supposed to be a baseline and we get reasonable denseclus results. I'm open either way here, any thoughts?
  • in the select_dtypes line why are we dropping segment then adding it back in the next line?
  • can we expand the analysis to look at more than just mean? I think other descriptive stats might help with the story telling (but understand cluster quality is not good so there isnt much of a story to tell)
  • can we see the columns used for categorical clustering?
  • categorical analysis points 2 and 3 seem to be conflicting: we are saying both that there is a small finite space where we can have points + we have a large sparse space

Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

can we update all of the plots to have appropriate x/y axis labels (or remove the labels) instead of None

momonga-ml marked this conversation as resolved.
Show resolved Hide resolved
momonga-ml marked this conversation as resolved.
Show resolved Hide resolved
momonga-ml marked this conversation as resolved.
Show resolved Hide resolved
momonga-ml marked this conversation as resolved.
Show resolved Hide resolved

Large diffs are not rendered by default.

4 changes: 2 additions & 2 deletions requirements-dev.txt
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,7 +1,7 @@
black==23.11.0
coverage==7.3.2
mypy==1.7.0
nbqa==1.7.0
mypy==1.7.1
nbqa==1.7.1
pre-commit==3.5.0
pylint==3.0.2
pytest==7.4.3
Expand Down
3 changes: 2 additions & 1 deletion requirements.txt
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -3,4 +3,5 @@ numpy>=1.20.2
hdbscan>=0.8.27
numba>=0.51.2
pandas>=1.2.4
scikit_learn>=0.24.2
scikit_learn>=0.24.2
seaborn>=0.13.0
Loading