fitting error

[mhady7]

I get "ValueError: Input contains NaN, infinity or a value too large for dtype('float64')" when trying to fit the data . Can you please help?

In [17]: b=np.random.random((300, 7000))

In [18]: b
Out[18]:
array([[ 0.32307771, 0.25185051, 0.65950904, ..., 0.07361776,
0.14508455, 0.95536933],
[ 0.73879883, 0.15227882, 0.73709136, ..., 0.46127292,
0.16149633, 0.55478285],
[ 0.33343231, 0.74376996, 0.17680955, ..., 0.17281043,
0.69592014, 0.11823333],
...,
[ 0.12729313, 0.70065767, 0.16744481, ..., 0.97365921,
0.32745413, 0.82913982],
[ 0.57948363, 0.8199712 , 0.40820909, ..., 0.36342576,
0.67581305, 0.80445348],
[ 0.75098905, 0.56323658, 0.91718958, ..., 0.86885992,
0.81540044, 0.38880498]])

In [19]: a.fit(b)
/share/apps/python/2.7.11/Canopy_64bit/User/lib/python2.7/site-packages/sklearn/utils/validation.py:386: DeprecationWarning: Passing 1d arrays as data is deprecated in 0.17 and willraise ValueError in 0.19. Reshape your data either using X.reshape(-1, 1) if your data has a single feature or X.reshape(1, -1) if it contains a single sample.
DeprecationWarning)
/share/apps/python/2.7.11/Canopy_64bit/User/lib/python2.7/site-packages/sklearn/utils/validation.py:386: DeprecationWarning: Passing 1d arrays as data is deprecated in 0.17 and willraise ValueError in 0.19. Reshape your data either using X.reshape(-1, 1) if your data has a single feature or X.reshape(1, -1) if it contains a single sample.
DeprecationWarning)

ValueError Traceback (most recent call last)
in ()
----> 1 a.fit(b)

/scratch/mohamed.ayed/python/lib/python2.7/site-packages/pyxmeans/xmeans.pyc in fit(self, data)
65
66 bic_parent = XMeans.bic([points], [centroid,])
---> 67 bic_child = XMeans.bic([cluster1, cluster2], test_model.centroids)
68 logging.info("\t\tbic_parent = %f, bic_child = %f", bic_parent, bic_child)
69 if bic_child > bic_parent:

/scratch/mohamed.ayed/python/lib/python2.7/site-packages/pyxmeans/xmeans.pyc in bic(cls, clusters, centroids)
97 num_dims = clusters[0][0].shape[0]
98
---> 99 log_likelihood = XMeans._loglikelihood(num_points, num_dims, clusters, centroids)
100 num_params = XMeans._free_params(len(clusters), num_dims)
101

/scratch/mohamed.ayed/python/lib/python2.7/site-packages/pyxmeans/xmeans.pyc in _loglikelihood(cls, num_points, num_dims, clusters, centroids)
115 t1 = fRn * np.log(fRn)
116 t2 = fRn * np.log(num_points)
--> 117 variance = XMeans._cluster_variance(num_points, clusters, centroids) or np.nextafter(0, 1)
118 t3 = ((fRn * num_dims) / 2.0) * np.log((2.0 * np.pi) * variance)
119 t4 = (fRn - 1.0) / 2.0

/scratch/mohamed.ayed/python/lib/python2.7/site-packages/pyxmeans/xmeans.pyc in _cluster_variance(cls, num_points, clusters, centroids)
127 denom = float(num_points - len(centroids))
128 for cluster, centroid in zip(clusters, centroids):
--> 129 distances = euclidean_distances(cluster, centroid)
130 s += (distances*distances).sum()
131 return s / denom

/share/apps/python/2.7.11/Canopy_64bit/User/lib/python2.7/site-packages/sklearn/metrics/pairwise.py in euclidean_distances(X, Y, Y_norm_squared, squared, X_norm_squared)
206 paired_distances : distances betweens pairs of elements of X and Y.
207 """
--> 208 X, Y = check_pairwise_arrays(X, Y)
209
210 if X_norm_squared is not None:

/share/apps/python/2.7.11/Canopy_64bit/User/lib/python2.7/site-packages/sklearn/metrics/pairwise.py in check_pairwise_arrays(X, Y, precomputed)
95 else:
96 X = check_array(X, accept_sparse='csr', dtype=dtype)
---> 97 Y = check_array(Y, accept_sparse='csr', dtype=dtype)
98
99 if precomputed:

/share/apps/python/2.7.11/Canopy_64bit/User/lib/python2.7/site-packages/sklearn/utils/validation.py in check_array(array, accept_sparse, dtype, order, copy, force_all_finite, ensure_2d, allow_nd, ensure_min_samples, ensure_min_features, warn_on_dtype, estimator)
396 % (array.ndim, estimator_name))
397 if force_all_finite:
--> 398 _assert_all_finite(array)
399
400 shape_repr = _shape_repr(array.shape)

/share/apps/python/2.7.11/Canopy_64bit/User/lib/python2.7/site-packages/sklearn/utils/validation.py in _assert_all_finite(X)
52 and not np.isfinite(X).all()):
53 raise ValueError("Input contains NaN, infinity"
---> 54 " or a value too large for %r." % X.dtype)
55
56

ValueError: Input contains NaN, infinity or a value too large for dtype('float64').