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

feat(dendrogram): add customizable metric parameter for distance calculation #174

Open
wants to merge 1 commit into
base: master
Choose a base branch
from
Open

feat(dendrogram): add customizable metric parameter for distance calculation #174

Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
23 changes: 18 additions & 5 deletions missingno/missingno.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -413,19 +413,32 @@ def heatmap(


def dendrogram(
df, method='average', filter=None, n=0, p=0, orientation=None, figsize=None, fontsize=16,
label_rotation=45, ax=None
df, method='average', metric='euclidean', filter=None, n=0, p=0, orientation=None,
figsize=None, fontsize=16, label_rotation=45, ax=None
):
"""
Fits a `scipy` hierarchical clustering algorithm to the given DataFrame's variables and visualizes the results as
a `scipy` dendrogram.

The default vertical display will fit up to 50 columns. If more than 50 columns are specified and orientation is
left unspecified the dendrogram will automatically swap to a horizontal display to fit the additional variables.

:param df: The DataFrame whose completeness is being dendrogrammed.
:param method: The distance measure being used for clustering. This is a parameter that is passed to
`scipy.hierarchy`.

:param metric: The distance metric for clustering. Defaults to "euclidean".
Available options include:
- "euclidean": Standard geometric distance, commonly used for continuous data.
- "cityblock" (or "manhattan"): Sum of absolute differences, suitable for high-dimensional data.
- "cosine": Cosine of the angle between two vectors, ideal for text or high-dimensional data.
- "hamming": Proportion of differing elements, recommended for binary data (e.g., presence/absence).
- "jaccard": Ratio of intersection to union, also recommended for binary or sparse data.
- "chebyshev": Maximum difference along any coordinate dimension.
- "minkowski": Generalized distance metric (parameterized by `p`); Euclidean if `p=2`, Manhattan if `p=1`.
- "braycurtis": Dissimilarity measure, often used in ecological data comparisons.
- "canberra": Sum of ratio differences; sensitive to small changes, useful for environmental data.
- "correlation": 1 minus the Pearson correlation; good for continuous data and pattern similarity.

:param filter: The filter to apply to the heatmap. Should be one of "top", "bottom", or None (default).
:param n: The cap on the number of columns to include in the filtered DataFrame.
:param p: The cap on the percentage fill of the columns in the filtered DataFrame.
Expand All @@ -450,9 +463,9 @@ def dendrogram(

df = nullity_filter(df, filter=filter, n=n, p=p)

# Link the hierarchical output matrix, figure out orientation, construct base dendrogram.
# NOTE: A new `metric` parameter allows for alternate distance measures such as "hamming" or "jaccard"
x = np.transpose(df.isnull().astype(int).values)
z = hierarchy.linkage(x, method)
z = hierarchy.linkage(x, method=method, metric=metric)

if not orientation:
if len(df.columns) > 50:
Expand Down