We Keep Coding

Is it possible to find eigenvalues (and vectors) using Dask, and if so, how?

I am only just starting to look into Dask. I think it might be practical for a script which today uses numpy functions, but which I guess could be "upgraded" for solving even bigger problems by using Dask. So pretty much the "center" of the script is to find eigenvalues and -vectors. This is done via np.linalg.eigh today.
It seems that right now, Dask does not have a da.linalg.eigh function. Is there a way to achieve this through other (dask) means, or do I have to stick with the previous numpy approach?

index milion row square matrix for fast access

I have some very large matrices (let say of the order of the million rows), that I can not keep in memory, and I would need to access to subsample of this matrix in descent time (less than a minute...).
I started looking at hdf5 and blaze in combination with numpy and pandas:
http://web.datapark.io/yves/blaze.html
http://blaze.pydata.org
But I found it a bit complicated, and I am not sure if it is the best solution.
Are there other solutions?
thanks
EDIT
Here some more specifications about the kind of data I am dealing with.
The matrices are usually sparse (< 10% or < 25% of cells with non-zero)
The matrices are symmetric
And what I would need to do is:
Access for reading only
Extract rectangular sub-matrices (mostly along the diagonal, but also outside)

Did you try PyTables ? It can be very useful for very large matrix. Take a look to this SO post.

Your question is lacking a bit in context; but hdf5 compressed block storage is probably as-efficient as a sparse storage format for these relatively dense matrices you describe. In memory, you can always cast your views to sparse matrices if it pays. That seems like an effective and simple solution; and as far as I know there are no sparse matrix formats which can easily be read partially from disk.

Python - memory usage

I am a computer engineer student and I started to work with Python. Now, my assignment is create an matrix but very large scale. How can I handle it in order to take less memory? I did some search and found "memory handler", but I cant be sure if the handler can bu used for this. Or are there any module in Python library?
Thank you.

You should be looking into numpy and scipy. They are relatively thin layers on top of blocks of memory, and are usually quite efficient for matrix type calculations. If your matrix is large but sparse (ie most elements are 0), have a look at scipy's sparse matrices.

Best module to do large matrix computation in Python?

I am developing a simple recommendation system and trying to do some computation like SVD, RBM, etc.
To be more convincing, I am going to use the Movielens or Netflix dataset to evaluate the performance of the system. However, the two datasets both have more than 1 million of users and more than 10 thousand of items, it's impossible to put all the data into memory. I have to use some specific modules to handle such a large matrix.
I know there are some tools in SciPy can handle this, and divisi2 used by python-recsys also seems like a good choice. Or maybe there are some better tools I don't know?
Which module should I use? Any suggestion?

I would suggest SciPy, specifically Sparse. As Dougal pointed out, Numpy is not suited for this situation.

I found another solution named crab, I try finding and comparing some of them.

If your concern is just putting the data in the memory use 64bit python with 64bit numpy. If you dont have enough physical memory you can just increase virtual memory in os level. The size of virtual memory is only limited by your hdd size. Speed of computation however is a different beast!

Automatic CudaMat conversion in Python

I'm looking into speeding up my python code, which is all matrix math, using some form of CUDA. Currently my code is using Python and Numpy, so it seems like it shouldn't be too difficult to rewrite it using something like either PyCUDA or CudaMat.
However, on my first attempt using CudaMat, I realized I had to rearrange a lot of the equations in order to keep the operations all on the GPU. This included the creation of many temporary variables so I could store the results of the operations.
I understand why this is necessary, but it makes what were once easy to read equations into somewhat of a mess that difficult to inspect for correctness. Additionally, I would like to be able to easily modify the equations later on, which isn't in their converted form.
The package Theano manages to do this by first creating a symbolic representation of the operations, then compiling them to CUDA. However, after trying Theano out for a bit, I was frustrated by how opaque everything was. For example, just getting the actual value for myvar.shape[0] is made difficult since the tree doesn't get evaluated until much later. I would also much prefer less of a framework in which my code much conform to a library that acts invisibly in the place of Numpy.
Thus, what I would really like is something much simpler. I don't want automatic differentiation (there are other packages like OpenOpt that can do that if I require it), or optimization of the tree, but just a conversion from standard Numpy notation to CudaMat/PyCUDA/somethingCUDA. In fact, I want to be able to have it evaluate to just Numpy without any CUDA code for testing.
I'm currently considering writing this myself, but before even consider such a venture, I wanted to see if anyone else knows of similar projects or a good starting place. The only other project I know that might be close to this is SymPy, but I don't know how easy it would be to adapt to this purpose.
My current idea would be to create an array class that looked like a Numpy.array class. It's only function would be to build a tree. At any time, that symbolic array class could be converted to a Numpy array class and be evaluated (there would also be a one-to-one parity). Alternatively, the array class could be traversed and have CudaMat commands be generated. If optimizations are required they can be done at that stage (e.g. re-ordering of operations, creation of temporary variables, etc.) without getting in the way of inspecting what's going on.
Any thoughts/comments/etc. on this would be greatly appreciated!
Update
A usage case may look something like (where sym is the theoretical module), where we might be doing something such as calculating the gradient:
W = sym.array(np.rand(size=(numVisible, numHidden)))
delta_o = -(x - z)
delta_h = sym.dot(delta_o, W)*h*(1.0-h)
grad_W = sym.dot(X.T, delta_h)
In this case, grad_W would actually just be a tree containing the operations that needed to be done. If you wanted to evaluate the expression normally (i.e. via Numpy) you could do:
npGrad_W = grad_W.asNumpy()
which would just execute the Numpy commands that the tree represents. If on the other hand, you wanted to use CUDA, you would do:
cudaGrad_W = grad_W.asCUDA()
which would convert the tree into expressions that can executed via CUDA (this could happen in a couple of different ways).
That way it should be trivial to: (1) test grad_W.asNumpy() == grad_W.asCUDA(), and (2) convert your pre-existing code to use CUDA.

Have you looked at the GPUArray portion of PyCUDA?
http://documen.tician.de/pycuda/array.html
While I haven't used it myself, it seems like it would be what you're looking for. In particular, check out the "Single-pass Custom Expression Evaluation" section near the bottom of that page.