I would like to define a covariance matrix in RStan.
Similarly to how you can provide constraints to scalar and vector values, e.g. real a, I would like to provide constraints that the leading diagonal of the covariance matrix must be positive, but the off-diagonal components could take any real value.
Is there a way to enforce that the matrix must also be positive semi-definite? Otherwise, some of the samples produced will not be valid covariance matrices.
Yes, defining
cov_matrix[K] Sigma;
ensures that Sigma is symmetric and positive definite K x K matrix. It can reduce to semidefinite due to floating point, but we'll catch that and raise exceptions to ensure it stays strictly positive definite.
Under the hood, Stan uses the Cholesky factor transform---the unconstrained representation is a lower triangular matrix with positive diagonal. We just use that as the real parameters, then transform and apply the Jacobian implicitly under the hood as described the reference manual chapter on constrained variables to create a covariance matrix with an implicit (improper) uniform prior.
Related
Now i have a 7000*7000 correlation matrix and I have to do PCA on this in R.
I used the
CorPCA <- princomp(covmat=xCor)
, xCor is the correlation matrix
but it comes out
"covariance matrix is not non-negative definite"
it is because i have some negative correlation in that matrix.
I am wondering which inbuilt function in R that i can use to get the result of PCA
One method to do the PCA is to perform an eigenvalue decomposition of the covariance matrix, see wikipedia.
The advantage of the eigenvalue decomposition is that you see which directions (eigenvectors) are significant, i.e. have a noticeable variation expressed by the associated eigenvalues. Moreover, you can detect if the covariance matrix is positive definite (all eigenvalues greater than zero), not negative-definite (which is okay) if there are eigenvalues equal zero or if it is indefinite (which is not okay) by negative eigenvalues. Sometimes it also happens that due to numerical inaccuracies a non-negative-definite matrix becomes negative-definite. In that case you would observe negative eigenvalues which are almost zero. In that case you can set these eigenvalues to zero to retain the non-negative definiteness of the covariance matrix. Furthermore, you can still interpret the result: eigenvectors contributing the significant information are associated with the biggest eigenvalues. If the list of sorted eigenvalues declines quickly there are a lot of directions which do not contribute significantly and therefore can be dropped.
The built-in R function is eigen
If your covariance matrix is A then
eigen_res <- eigen(A)
# sorted list of eigenvalues
eigen_res$values
# slightly negative eigenvalues, set them to small positive value
eigen_res$values[eigen_res$values<0] <- 1e-10
# and produce regularized covariance matrix
Areg <- eigen_res$vectors %*% diag(eigen_res$values) %*% t(eigen_res$vectors)
not non-negative definite does not mean the covariance matrix has negative correlations. It's a linear algebra equivalent of trying to take square root of negative number! You can't tell by looking at a few values of the matrix, whether it's positive definite.
Try adjusting some default values like tolerance in princomp call. Check this thread for example: How to use princomp () function in R when covariance matrix has zero's?
An alternative is to write some code of your own to perform what is called a n NIPLAS analysis. Take a look at this thread on the R-mailing list: https://stat.ethz.ch/pipermail/r-help/2006-July/110035.html
I'd even go as far as asking where did you obtain the correlation matrix? Did you construct it yourself? Does it have NAs? If you constructed xCor from your own data, do you think you can sample the data and construct a smaller xCor matrix? (say 1000X1000). All these alternatives try to drive your PCA algorithm through the 'happy path' (i.e. all matrix operations can be internally carried out without difficulties in diagonalization etc..i.e., no more 'non-negative definite error msgs)
I am doing a project which requires me to normalize a sparse NxNmatrix. I read somewhere that we can normalize a matrix so that its eigen values lie between [-1,1] by multiplying it with a diagonal matrix D such that N = D^{-1/2}*A*D^{-1/2}.
But I am not sure what D is here. Also, is there a function in Matlab that can do this normalization for sparse matrices?
It's possible that I am misunderstanding your question, but as it reads it makes no sense to me.
A matrix is just a representation of a linear transformation. Given that a matrix A corresponds to a linear transformation T, any matrix of the form B^{-1} A B (called the conjugate of A by B) for an invertible matrix B corresponds to the same transformation, represented in a difference basis. In particular, the eigen values of a matrix correspond to the eigen values of the linear transformation, so conjugating by an invertible matrix cannot change the eigen values.
It's possible that you meant that you want to scale the eigen vectors so that each has unit length. This is a common thing to do since then the eigen values tell you how far a vector of unit length is magnified by the transformation.
I have two square matrices A and B. A is symmetric, B is symmetric positive definite. I would like to compute $trace(A.B^{-1})$. For now, I compute the Cholesky decomposition of B, solve for C in the equation $A=C.B$ and sum up the diagonal elements.
Is there a more efficient way of proceeding?
I plan on using Eigen. Could you provide an implementation if the matrices are sparse (A can often be diagonal, B is often band-diagonal)?
If B is sparse, it may be efficient (i.e., O(n), assuming good condition number of B) to solve for x_i in
B x_i = a_i
(sample Conjugate Gradient code is given on Wikipedia). Taking a_i to be the column vectors of A, you get the matrix B^{-1} A in O(n^2). Then you can sum the diagonal elements to get the trace. Generally, it's easier to do this sparse inverse multiplication than to get the full set of eigenvalues. For comparison, Cholesky decomposition is O(n^3). (see Darren Engwirda's comment below about Cholesky).
If you only need an approximation to the trace, you can actually reduce the cost to O(q n) by averaging
r^T (A B^{-1}) r
over q random vectors r. Usually q << n. This is an unbiased estimate provided that the components of the random vector r satisfy
< r_i r_j > = \delta_{ij}
where < ... > indicates an average over the distribution of r. For example, components r_i could be independent gaussian distributed with unit variance. Or they could be selected uniformly from +-1. Typically the trace scales like O(n) and the error in the trace estimate scales like O(sqrt(n/q)), so the relative error scales as O(sqrt(1/nq)).
If generalized eigenvalues are more efficient to compute, you can compute the generalized eigenvalues, A*v = lambda* B *v and then sum up all the lambdas.
How would you generate a random matrix that is not singular in MATLAB.?
I know a matrix determinant can be used to do check this, but after reading MATLAB determinant:
"Using det(X) == 0 as a test for matrix singularity is appropriate only for matrices of modest order with small integer entries. Testing singularity using abs(det(X)) <= tolerance is not recommended as it is difficult to choose the correct tolerance. The function cond(X) can check for singular and nearly singular matrices."
So if I want to generate a big random matrix (axb) a=5000, b=5000, How to do it??.
A randomly generated matrix will be full rank (and hence invertible, if square) with probability 1:
A = randn(5000);
you can check this by using min(svd(A)), and verifying that the smallest singular value is larger than zero.
This is a well-known fact, but here's an example paper if you want one.
How does it actually reduce noise..can you suggest some nice tutorials?
SVD can be understood from a geometric sense for square matrices as a transformation on a vector.
Consider a square n x n matrix M multiplying a vector v to produce an output vector w:
w = M*v
The singular value decomposition M is the product of three matrices M=U*S*V, so w=U*S*V*v. U and V are orthonormal matrices. From a geometric transformation point of view (acting upon a vector by multiplying it), they are combinations of rotations and reflections that do not change the length of the vector they are multiplying. S is a diagonal matrix which represents scaling or squashing with different scaling factors (the diagonal terms) along each of the n axes.
So the effect of left-multiplying a vector v by a matrix M is to rotate/reflect v by M's orthonormal factor V, then scale/squash the result by a diagonal factor S, then rotate/reflect the result by M's orthonormal factor U.
One reason SVD is desirable from a numerical standpoint is that multiplication by orthonormal matrices is an invertible and extremely stable operation (condition number is 1). SVD captures any ill-conditioned-ness in the diagonal scaling matrix S.
One way to use SVD to reduce noise is to do the decomposition, set components that are near zero to be exactly zero, then re-compose.
Here's an online tutorial on SVD.
You might want to take a look at Numerical Recipes.
Singular value decomposition is a method for taking an nxm matrix M and "decomposing" it into three matrices such that M=USV. S is a diagonal square (the only nonzero entries are on the diagonal from top-left to bottom-right) matrix containing the "singular values" of M. U and V are orthogonal, which leads to the geometric understanding of SVD, but that isn't necessary for noise reduction.
With M=USV, we still have the original matrix M with all its noise intact. However, if we only keep the k largest singular values (which is easy, since many SVD algorithms compute a decomposition where the entries of S are sorted in nonincreasing order), then we have an approximation of the original matrix. This works because we assume that the small values are the noise, and that the more significant patterns in the data will be expressed through the vectors associated with larger singular values.
In fact, the resulting approximation is the most accurate rank-k approximation of the original matrix (has the least squared error).
To answer to the tittle question: SVD is a generalization of eigenvalues/eigenvectors to non-square matrices.
Say,
$X \in N \times p$, then the SVD decomposition of X yields X=UDV^T where D is diagonal and U and V are orthogonal matrices.
Now X^TX is a square matrice, and the SVD decomposition of X^TX=VD^2V where V is equivalent to the eigenvectors of X^TX and D^2 contains the eigenvalues of X^TX.
SVD can also be used to greatly ease global (i.e. to all observations simultaneously) fitting of an arbitrary model (expressed in an formula) to data (with respect to two variables and expressed in a matrix).
For example, data matrix A = D * MT where D represents the possible states of a system and M represents its evolution wrt some variable (e.g. time).
By SVD, A(x,y) = U(x) * S * VT(y) and therefore D * MT = U * S * VT
then D = U * S * VT * MT+ where the "+" indicates a pseudoinverse.
One can then take a mathematical model for the evolution and fit it to the columns of V, each of which are a linear combination the components of the model (this is easy, as each column is a 1D curve). This obtains model parameters which generate M? (the ? indicates it is based on fitting).
M * M?+ * V = V? which allows residuals R * S2 = V - V? to be minimized, thus determining D and M.
Pretty cool, eh?
The columns of U and V can also be inspected to glean information about the data; for example each inflection point in the columns of V typically indicates a different component of the model.
Finally, and actually addressing your question, it is import to note that although each successive singular value (element of the diagonal matrix S) with its attendant vectors U and V does have lower signal to noise, the separation of the components of the model in these "less important" vectors is actually more pronounced. In other words, if the data is described by a bunch of state changes that follow a sum of exponentials or whatever, the relative weights of each exponential get closer together in the smaller singular values. In other other words the later singular values have vectors which are less smooth (noisier) but in which the change represented by each component are more distinct.