Last time I checked what I want to do was not possible using sympy but several versions have come out since then, adding a few features that I just can't get to work.
First, what I am already doing is generating numpy arrays containing (large) sympy expressions in each element. Currently I am evaluating these homemade tensor expressions by iterating over the ndarray and using eval() for each expression and to get a numeric version. This is very slow, for obvious reasons. I would like to generate C code that evaluates these large expressions for me. In theory, I could have a python script generate a C file that contains one function for each element of my ndarray and then another, handwritten C file that calls each of these functions in a loop and populates and C array for me. However, I happened to be reading this page here http://docs.sympy.org/latest/modules/utilities/autowrap.html and the first example is an autowrap of a tensor operation, so I tried combining this with the codegen module and it worked like a charm. The problem is, my expressions require more complicated operations, like taking the derivative of a tensor with respect to another tensor, and I have no clue how to use the tensor module for this. Is it even possible, or is code generation for tensors limited to Einstein summation for now? -- You received this message because you are subscribed to the Google Groups "sympy" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To post to this group, send email to [email protected]. Visit this group at http://groups.google.com/group/sympy. For more options, visit https://groups.google.com/groups/opt_out.
