0

I'm trying to use UMAP in Jupyter (on Anaconda's environment) but something is not working properly. Below is the code I used.

# INSTALL UMAP
conda install -c conda-forge umap-learn    # this was what was mentioned on umap's website
import umap.umap_ as umap
reducer = umap.UMAP()

# DROP NON-NUMERIC DATA
a_shortData = shortData.drop(columns='Severity')

# CONVERT EACH FEATURE INTO Z-SCORES
scaled_data = StandardScaler().fit_transform(a_shortData)

# COMPUTES MEAN & STD OF DATA AND USE THESE TO AUTOSCALE THE DATA
embedding = reducer.fit_transform(scaled_data)
embedding.shape

And the error message is as below.

enter image description here enter image description here

Lucy
  • 17
  • 7
  • Please add code and data as text ([using code formatting](//stackoverflow.com/editing-help#code)), not images. Images: A) don't allow us to copy-&-paste the code/errors/data for testing; B) don't permit searching based on the code/error/data contents; and [many more reasons](//meta.stackoverflow.com/a/285557). Images should only be used, in addition to text in code format, if having the image adds something significant that is not conveyed by just the text code/error/data. –  Dec 11 '21 at 01:20
  • If you would like more specific help with this, consider adding information about the Conda environment, e.g., either the full environment YAML (`conda env export`) or a subset of pertinent packages (`conda list '(umap|numpy|numba|pandas)'`). Also, a minimal data example that can reproduce the error would also be helpful. – merv Dec 13 '21 at 21:25

0 Answers0