""" Extended math utilities. """ # Authors: Gael Varoquaux # Alexandre Gramfort # Alexandre T. Passos # Olivier Grisel # Lars Buitinck # Stefan van der Walt # Kyle Kastner # Giorgio Patrini # License: BSD 3 clause from __future__ import division from functools import partial import warnings import numpy as np from scipy import linalg from scipy.sparse import issparse, csr_matrix from . import check_random_state from .fixes import np_version from ._logistic_sigmoid import _log_logistic_sigmoid from ..externals.six.moves import xrange from .sparsefuncs_fast import csr_row_norms from .validation import check_array from ..exceptions import NonBLASDotWarning def norm(x): """Compute the Euclidean or Frobenius norm of x. Returns the Euclidean norm when x is a vector, the Frobenius norm when x is a matrix (2-d array). More precise than sqrt(squared_norm(x)). """ x = np.asarray(x) nrm2, = linalg.get_blas_funcs(['nrm2'], [x]) return nrm2(x) # Newer NumPy has a ravel that needs less copying. if np_version < (1, 7, 1): _ravel = np.ravel else: _ravel = partial(np.ravel, order='K') def squared_norm(x): """Squared Euclidean or Frobenius norm of x. Returns the Euclidean norm when x is a vector, the Frobenius norm when x is a matrix (2-d array). Faster than norm(x) ** 2. """ x = _ravel(x) return np.dot(x, x) def row_norms(X, squared=False): """Row-wise (squared) Euclidean norm of X. Equivalent to np.sqrt((X * X).sum(axis=1)), but also supports sparse matrices and does not create an X.shape-sized temporary. Performs no input validation. """ if issparse(X): if not isinstance(X, csr_matrix): X = csr_matrix(X) norms = csr_row_norms(X) else: norms = np.einsum('ij,ij->i', X, X) if not squared: np.sqrt(norms, norms) return norms def fast_logdet(A): """Compute log(det(A)) for A symmetric Equivalent to : np.log(nl.det(A)) but more robust. It returns -Inf if det(A) is non positive or is not defined. """ sign, ld = np.linalg.slogdet(A) if not sign > 0: return -np.inf return ld def _impose_f_order(X): """Helper Function""" # important to access flags instead of calling np.isfortran, # this catches corner cases. if X.flags.c_contiguous: return check_array(X.T, copy=False, order='F'), True else: return check_array(X, copy=False, order='F'), False def _fast_dot(A, B): if B.shape[0] != A.shape[A.ndim - 1]: # check adopted from '_dotblas.c' raise ValueError if A.dtype != B.dtype or any(x.dtype not in (np.float32, np.float64) for x in [A, B]): warnings.warn('Falling back to np.dot. ' 'Data must be of same type of either ' '32 or 64 bit float for the BLAS function, gemm, to be ' 'used for an efficient dot operation. ', NonBLASDotWarning) raise ValueError if min(A.shape) == 1 or min(B.shape) == 1 or A.ndim != 2 or B.ndim != 2: raise ValueError # scipy 0.9 compliant API dot = linalg.get_blas_funcs(['gemm'], (A, B))[0] A, trans_a = _impose_f_order(A) B, trans_b = _impose_f_order(B) return dot(alpha=1.0, a=A, b=B, trans_a=trans_a, trans_b=trans_b) def _have_blas_gemm(): try: linalg.get_blas_funcs(['gemm']) return True except (AttributeError, ValueError): warnings.warn('Could not import BLAS, falling back to np.dot') return False # Only use fast_dot for older NumPy; newer ones have tackled the speed issue. if np_version < (1, 7, 2) and _have_blas_gemm(): def fast_dot(A, B): """Compute fast dot products directly calling BLAS. This function calls BLAS directly while warranting Fortran contiguity. This helps avoiding extra copies `np.dot` would have created. For details see section `Linear Algebra on large Arrays`: http://wiki.scipy.org/PerformanceTips Parameters ---------- A, B: instance of np.ndarray Input arrays. Arrays are supposed to be of the same dtype and to have exactly 2 dimensions. Currently only floats are supported. In case these requirements aren't met np.dot(A, B) is returned instead. To activate the related warning issued in this case execute the following lines of code: >> import warnings >> from sklearn.exceptions import NonBLASDotWarning >> warnings.simplefilter('always', NonBLASDotWarning) """ try: return _fast_dot(A, B) except ValueError: # Maltyped or malformed data. return np.dot(A, B) else: fast_dot = np.dot def density(w, **kwargs): """Compute density of a sparse vector Return a value between 0 and 1 """ if hasattr(w, "toarray"): d = float(w.nnz) / (w.shape[0] * w.shape[1]) else: d = 0 if w is None else float((w != 0).sum()) / w.size return d def safe_sparse_dot(a, b, dense_output=False): """Dot product that handle the sparse matrix case correctly Uses BLAS GEMM as replacement for numpy.dot where possible to avoid unnecessary copies. """ if issparse(a) or issparse(b): ret = a * b if dense_output and hasattr(ret, "toarray"): ret = ret.toarray() return ret else: return fast_dot(a, b) def randomized_range_finder(A, size, n_iter, power_iteration_normalizer='auto', random_state=None): """Computes an orthonormal matrix whose range approximates the range of A. Parameters ---------- A: 2D array The input data matrix size: integer Size of the return array n_iter: integer Number of power iterations used to stabilize the result power_iteration_normalizer: 'auto' (default), 'QR', 'LU', 'none' Whether the power iterations are normalized with step-by-step QR factorization (the slowest but most accurate), 'none' (the fastest but numerically unstable when `n_iter` is large, e.g. typically 5 or larger), or 'LU' factorization (numerically stable but can lose slightly in accuracy). The 'auto' mode applies no normalization if `n_iter`<=2 and switches to LU otherwise. .. versionadded:: 0.18 random_state: RandomState or an int seed (0 by default) A random number generator instance Returns ------- Q: 2D array A (size x size) projection matrix, the range of which approximates well the range of the input matrix A. Notes ----- Follows Algorithm 4.3 of Finding structure with randomness: Stochastic algorithms for constructing approximate matrix decompositions Halko, et al., 2009 (arXiv:909) http://arxiv.org/pdf/0909.4061 An implementation of a randomized algorithm for principal component analysis A. Szlam et al. 2014 """ random_state = check_random_state(random_state) # Generating normal random vectors with shape: (A.shape[1], size) Q = random_state.normal(size=(A.shape[1], size)) # Deal with "auto" mode if power_iteration_normalizer == 'auto': if n_iter <= 2: power_iteration_normalizer = 'none' else: power_iteration_normalizer = 'LU' # Perform power iterations with Q to further 'imprint' the top # singular vectors of A in Q for i in range(n_iter): if power_iteration_normalizer == 'none': Q = safe_sparse_dot(A, Q) Q = safe_sparse_dot(A.T, Q) elif power_iteration_normalizer == 'LU': Q, _ = linalg.lu(safe_sparse_dot(A, Q), permute_l=True) Q, _ = linalg.lu(safe_sparse_dot(A.T, Q), permute_l=True) elif power_iteration_normalizer == 'QR': Q, _ = linalg.qr(safe_sparse_dot(A, Q), mode='economic') Q, _ = linalg.qr(safe_sparse_dot(A.T, Q), mode='economic') # Sample the range of A using by linear projection of Q # Extract an orthonormal basis Q, _ = linalg.qr(safe_sparse_dot(A, Q), mode='economic') return Q def randomized_svd(M, n_components, n_oversamples=10, n_iter='auto', power_iteration_normalizer='auto', transpose='auto', flip_sign=True, random_state=0): """Computes a truncated randomized SVD Parameters ---------- M: ndarray or sparse matrix Matrix to decompose n_components: int Number of singular values and vectors to extract. n_oversamples: int (default is 10) Additional number of random vectors to sample the range of M so as to ensure proper conditioning. The total number of random vectors used to find the range of M is n_components + n_oversamples. Smaller number can improve speed but can negatively impact the quality of approximation of singular vectors and singular values. n_iter: int or 'auto' (default is 'auto') Number of power iterations. It can be used to deal with very noisy problems. When 'auto', it is set to 4, unless `n_components` is small (< .1 * min(X.shape)) `n_iter` in which case is set to 7. This improves precision with few components. .. versionchanged:: 0.18 power_iteration_normalizer: 'auto' (default), 'QR', 'LU', 'none' Whether the power iterations are normalized with step-by-step QR factorization (the slowest but most accurate), 'none' (the fastest but numerically unstable when `n_iter` is large, e.g. typically 5 or larger), or 'LU' factorization (numerically stable but can lose slightly in accuracy). The 'auto' mode applies no normalization if `n_iter`<=2 and switches to LU otherwise. .. versionadded:: 0.18 transpose: True, False or 'auto' (default) Whether the algorithm should be applied to M.T instead of M. The result should approximately be the same. The 'auto' mode will trigger the transposition if M.shape[1] > M.shape[0] since this implementation of randomized SVD tend to be a little faster in that case. .. versionchanged:: 0.18 flip_sign: boolean, (True by default) The output of a singular value decomposition is only unique up to a permutation of the signs of the singular vectors. If `flip_sign` is set to `True`, the sign ambiguity is resolved by making the largest loadings for each component in the left singular vectors positive. random_state: RandomState or an int seed (0 by default) A random number generator instance to make behavior Notes ----- This algorithm finds a (usually very good) approximate truncated singular value decomposition using randomization to speed up the computations. It is particularly fast on large matrices on which you wish to extract only a small number of components. In order to obtain further speed up, `n_iter` can be set <=2 (at the cost of loss of precision). References ---------- * Finding structure with randomness: Stochastic algorithms for constructing approximate matrix decompositions Halko, et al., 2009 http://arxiv.org/abs/arXiv:0909.4061 * A randomized algorithm for the decomposition of matrices Per-Gunnar Martinsson, Vladimir Rokhlin and Mark Tygert * An implementation of a randomized algorithm for principal component analysis A. Szlam et al. 2014 """ random_state = check_random_state(random_state) n_random = n_components + n_oversamples n_samples, n_features = M.shape if n_iter == 'auto': # Checks if the number of iterations is explicitely specified # Adjust n_iter. 7 was found a good compromise for PCA. See #5299 n_iter = 7 if n_components < .1 * min(M.shape) else 4 if transpose == 'auto': transpose = n_samples < n_features if transpose: # this implementation is a bit faster with smaller shape[1] M = M.T Q = randomized_range_finder(M, n_random, n_iter, power_iteration_normalizer, random_state) # project M to the (k + p) dimensional space using the basis vectors B = safe_sparse_dot(Q.T, M) # compute the SVD on the thin matrix: (k + p) wide Uhat, s, V = linalg.svd(B, full_matrices=False) del B U = np.dot(Q, Uhat) if flip_sign: if not transpose: U, V = svd_flip(U, V) else: # In case of transpose u_based_decision=false # to actually flip based on u and not v. U, V = svd_flip(U, V, u_based_decision=False) if transpose: # transpose back the results according to the input convention return V[:n_components, :].T, s[:n_components], U[:, :n_components].T else: return U[:, :n_components], s[:n_components], V[:n_components, :] def logsumexp(arr, axis=0): """Computes the sum of arr assuming arr is in the log domain. Returns log(sum(exp(arr))) while minimizing the possibility of over/underflow. Examples -------- >>> import numpy as np >>> from sklearn.utils.extmath import logsumexp >>> a = np.arange(10) >>> np.log(np.sum(np.exp(a))) 9.4586297444267107 >>> logsumexp(a) 9.4586297444267107 """ arr = np.rollaxis(arr, axis) # Use the max to normalize, as with the log this is what accumulates # the less errors vmax = arr.max(axis=0) out = np.log(np.sum(np.exp(arr - vmax), axis=0)) out += vmax return out def weighted_mode(a, w, axis=0): """Returns an array of the weighted modal (most common) value in a If there is more than one such value, only the first is returned. The bin-count for the modal bins is also returned. This is an extension of the algorithm in scipy.stats.mode. Parameters ---------- a : array_like n-dimensional array of which to find mode(s). w : array_like n-dimensional array of weights for each value axis : int, optional Axis along which to operate. Default is 0, i.e. the first axis. Returns ------- vals : ndarray Array of modal values. score : ndarray Array of weighted counts for each mode. Examples -------- >>> from sklearn.utils.extmath import weighted_mode >>> x = [4, 1, 4, 2, 4, 2] >>> weights = [1, 1, 1, 1, 1, 1] >>> weighted_mode(x, weights) (array([ 4.]), array([ 3.])) The value 4 appears three times: with uniform weights, the result is simply the mode of the distribution. >>> weights = [1, 3, 0.5, 1.5, 1, 2] # deweight the 4's >>> weighted_mode(x, weights) (array([ 2.]), array([ 3.5])) The value 2 has the highest score: it appears twice with weights of 1.5 and 2: the sum of these is 3. See Also -------- scipy.stats.mode """ if axis is None: a = np.ravel(a) w = np.ravel(w) axis = 0 else: a = np.asarray(a) w = np.asarray(w) axis = axis if a.shape != w.shape: w = np.zeros(a.shape, dtype=w.dtype) + w scores = np.unique(np.ravel(a)) # get ALL unique values testshape = list(a.shape) testshape[axis] = 1 oldmostfreq = np.zeros(testshape) oldcounts = np.zeros(testshape) for score in scores: template = np.zeros(a.shape) ind = (a == score) template[ind] = w[ind] counts = np.expand_dims(np.sum(template, axis), axis) mostfrequent = np.where(counts > oldcounts, score, oldmostfreq) oldcounts = np.maximum(counts, oldcounts) oldmostfreq = mostfrequent return mostfrequent, oldcounts def pinvh(a, cond=None, rcond=None, lower=True): """Compute the (Moore-Penrose) pseudo-inverse of a hermetian matrix. Calculate a generalized inverse of a symmetric matrix using its eigenvalue decomposition and including all 'large' eigenvalues. Parameters ---------- a : array, shape (N, N) Real symmetric or complex hermetian matrix to be pseudo-inverted cond : float or None, default None Cutoff for 'small' eigenvalues. Singular values smaller than rcond * largest_eigenvalue are considered zero. If None or -1, suitable machine precision is used. rcond : float or None, default None (deprecated) Cutoff for 'small' eigenvalues. Singular values smaller than rcond * largest_eigenvalue are considered zero. If None or -1, suitable machine precision is used. lower : boolean Whether the pertinent array data is taken from the lower or upper triangle of a. (Default: lower) Returns ------- B : array, shape (N, N) Raises ------ LinAlgError If eigenvalue does not converge Examples -------- >>> import numpy as np >>> a = np.random.randn(9, 6) >>> a = np.dot(a, a.T) >>> B = pinvh(a) >>> np.allclose(a, np.dot(a, np.dot(B, a))) True >>> np.allclose(B, np.dot(B, np.dot(a, B))) True """ a = np.asarray_chkfinite(a) s, u = linalg.eigh(a, lower=lower) if rcond is not None: cond = rcond if cond in [None, -1]: t = u.dtype.char.lower() factor = {'f': 1E3, 'd': 1E6} cond = factor[t] * np.finfo(t).eps # unlike svd case, eigh can lead to negative eigenvalues above_cutoff = (abs(s) > cond * np.max(abs(s))) psigma_diag = np.zeros_like(s) psigma_diag[above_cutoff] = 1.0 / s[above_cutoff] return np.dot(u * psigma_diag, np.conjugate(u).T) def cartesian(arrays, out=None): """Generate a cartesian product of input arrays. Parameters ---------- arrays : list of array-like 1-D arrays to form the cartesian product of. out : ndarray Array to place the cartesian product in. Returns ------- out : ndarray 2-D array of shape (M, len(arrays)) containing cartesian products formed of input arrays. Examples -------- >>> cartesian(([1, 2, 3], [4, 5], [6, 7])) array([[1, 4, 6], [1, 4, 7], [1, 5, 6], [1, 5, 7], [2, 4, 6], [2, 4, 7], [2, 5, 6], [2, 5, 7], [3, 4, 6], [3, 4, 7], [3, 5, 6], [3, 5, 7]]) """ arrays = [np.asarray(x) for x in arrays] shape = (len(x) for x in arrays) dtype = arrays[0].dtype ix = np.indices(shape) ix = ix.reshape(len(arrays), -1).T if out is None: out = np.empty_like(ix, dtype=dtype) for n, arr in enumerate(arrays): out[:, n] = arrays[n][ix[:, n]] return out def svd_flip(u, v, u_based_decision=True): """Sign correction to ensure deterministic output from SVD. Adjusts the columns of u and the rows of v such that the loadings in the columns in u that are largest in absolute value are always positive. Parameters ---------- u, v : ndarray u and v are the output of `linalg.svd` or `sklearn.utils.extmath.randomized_svd`, with matching inner dimensions so one can compute `np.dot(u * s, v)`. u_based_decision : boolean, (default=True) If True, use the columns of u as the basis for sign flipping. Otherwise, use the rows of v. The choice of which variable to base the decision on is generally algorithm dependent. Returns ------- u_adjusted, v_adjusted : arrays with the same dimensions as the input. """ if u_based_decision: # columns of u, rows of v max_abs_cols = np.argmax(np.abs(u), axis=0) signs = np.sign(u[max_abs_cols, xrange(u.shape[1])]) u *= signs v *= signs[:, np.newaxis] else: # rows of v, columns of u max_abs_rows = np.argmax(np.abs(v), axis=1) signs = np.sign(v[xrange(v.shape[0]), max_abs_rows]) u *= signs v *= signs[:, np.newaxis] return u, v def log_logistic(X, out=None): """Compute the log of the logistic function, ``log(1 / (1 + e ** -x))``. This implementation is numerically stable because it splits positive and negative values:: -log(1 + exp(-x_i)) if x_i > 0 x_i - log(1 + exp(x_i)) if x_i <= 0 For the ordinary logistic function, use ``sklearn.utils.fixes.expit``. Parameters ---------- X: array-like, shape (M, N) or (M, ) Argument to the logistic function out: array-like, shape: (M, N) or (M, ), optional: Preallocated output array. Returns ------- out: array, shape (M, N) or (M, ) Log of the logistic function evaluated at every point in x Notes ----- See the blog post describing this implementation: http://fa.bianp.net/blog/2013/numerical-optimizers-for-logistic-regression/ """ is_1d = X.ndim == 1 X = np.atleast_2d(X) X = check_array(X, dtype=np.float64) n_samples, n_features = X.shape if out is None: out = np.empty_like(X) _log_logistic_sigmoid(n_samples, n_features, X, out) if is_1d: return np.squeeze(out) return out def softmax(X, copy=True): """ Calculate the softmax function. The softmax function is calculated by np.exp(X) / np.sum(np.exp(X), axis=1) This will cause overflow when large values are exponentiated. Hence the largest value in each row is subtracted from each data point to prevent this. Parameters ---------- X: array-like, shape (M, N) Argument to the logistic function copy: bool, optional Copy X or not. Returns ------- out: array, shape (M, N) Softmax function evaluated at every point in x """ if copy: X = np.copy(X) max_prob = np.max(X, axis=1).reshape((-1, 1)) X -= max_prob np.exp(X, X) sum_prob = np.sum(X, axis=1).reshape((-1, 1)) X /= sum_prob return X def safe_min(X): """Returns the minimum value of a dense or a CSR/CSC matrix. Adapated from http://stackoverflow.com/q/13426580 """ if issparse(X): if len(X.data) == 0: return 0 m = X.data.min() return m if X.getnnz() == X.size else min(m, 0) else: return X.min() def make_nonnegative(X, min_value=0): """Ensure `X.min()` >= `min_value`.""" min_ = safe_min(X) if min_ < min_value: if issparse(X): raise ValueError("Cannot make the data matrix" " nonnegative because it is sparse." " Adding a value to every entry would" " make it no longer sparse.") X = X + (min_value - min_) return X def _incremental_mean_and_var(X, last_mean=.0, last_variance=None, last_sample_count=0): """Calculate mean update and a Youngs and Cramer variance update. last_mean and last_variance are statistics computed at the last step by the function. Both must be initialized to 0.0. In case no scaling is required last_variance can be None. The mean is always required and returned because necessary for the calculation of the variance. last_n_samples_seen is the number of samples encountered until now. From the paper "Algorithms for computing the sample variance: analysis and recommendations", by Chan, Golub, and LeVeque. Parameters ---------- X : array-like, shape (n_samples, n_features) Data to use for variance update last_mean : array-like, shape: (n_features,) last_variance : array-like, shape: (n_features,) last_sample_count : int Returns ------- updated_mean : array, shape (n_features,) updated_variance : array, shape (n_features,) If None, only mean is computed updated_sample_count : int References ---------- T. Chan, G. Golub, R. LeVeque. Algorithms for computing the sample variance: recommendations, The American Statistician, Vol. 37, No. 3, pp. 242-247 Also, see the sparse implementation of this in `utils.sparsefuncs.incr_mean_variance_axis` and `utils.sparsefuncs_fast.incr_mean_variance_axis0` """ # old = stats until now # new = the current increment # updated = the aggregated stats last_sum = last_mean * last_sample_count new_sum = X.sum(axis=0) new_sample_count = X.shape[0] updated_sample_count = last_sample_count + new_sample_count updated_mean = (last_sum + new_sum) / updated_sample_count if last_variance is None: updated_variance = None else: new_unnormalized_variance = X.var(axis=0) * new_sample_count if last_sample_count == 0: # Avoid division by 0 updated_unnormalized_variance = new_unnormalized_variance else: last_over_new_count = last_sample_count / new_sample_count last_unnormalized_variance = last_variance * last_sample_count updated_unnormalized_variance = ( last_unnormalized_variance + new_unnormalized_variance + last_over_new_count / updated_sample_count * (last_sum / last_over_new_count - new_sum) ** 2) updated_variance = updated_unnormalized_variance / updated_sample_count return updated_mean, updated_variance, updated_sample_count def _deterministic_vector_sign_flip(u): """Modify the sign of vectors for reproducibility Flips the sign of elements of all the vectors (rows of u) such that the absolute maximum element of each vector is positive. Parameters ---------- u : ndarray Array with vectors as its rows. Returns ------- u_flipped : ndarray with same shape as u Array with the sign flipped vectors as its rows. """ max_abs_rows = np.argmax(np.abs(u), axis=1) signs = np.sign(u[range(u.shape[0]), max_abs_rows]) u *= signs[:, np.newaxis] return u def stable_cumsum(arr, rtol=1e-05, atol=1e-08): """Use high precision for cumsum and check that final value matches sum Parameters ---------- arr : array-like To be cumulatively summed as flat rtol : float Relative tolerance, see ``np.allclose`` atol : float Absolute tolerance, see ``np.allclose`` """ out = np.cumsum(arr, dtype=np.float64) expected = np.sum(arr, dtype=np.float64) if not np.allclose(out[-1], expected, rtol=rtol, atol=atol): raise RuntimeError('cumsum was found to be unstable: ' 'its last element does not correspond to sum') return out