duhaime

import numpy as np
import umap

# generate some points in 4D
points = np.random.random((2000, 4))

# project the points to 2D to get points proximate to one another
points = umap.UMAP().fit_transform(points)

from lloyd.lloyd import Field

duhaime

These are specified in ~/.matplotlib/matplotlibrc

'svg', 'nbAgg', 'Qt4Cairo', 'pgf', 'Qt5Agg', 'TkCairo', 'WXAgg', 'template', 'GTK3Agg', 'pdf', 'Qt4Agg', 'MacOSX', 'WX', 'agg', 'Qt5Cairo', 'cairo', 'ps', 'WXCairo', 'GTK3Cairo', 'TkAgg', 'WebAgg'

duhaime

license: MIT
height: 500
scrolling: no
border: yes

duhaime

duhaime

license: MIT
height: 500
scrolling: no
border: yes

duhaime

license: MIT
height: 800
scrolling: no
border: yes

duhaime

import numpy as np
import sklearn.cluster
from distance import levenshtein

words = list(c.keys()) # list of words to cluster
words = np.asarray(words)
X = -1*np.array([[levenshtein(w1,w2) for w1 in words] for w2 in words])
clf = sklearn.cluster.AffinityPropagation(affinity="precomputed", damping=0.5)
clf.fit(X)

duhaime

license: gpl-3.0

duhaime

license: mit

duhaime