Hi, There is the Theano software in the python scientific tools that can do symbolic differentiation. With it, you can compute the gradient, the hessien and/or the jacobian. You can also compute efficiently jacobian times a vector, without explicitly computing the jacobian using R operation or L operation. http://deeplearning.net/software/theano/tutorial/gradients.html HTH Fred 2011/10/31 François Boulogne <boulogne.f@gmail.com>:
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
Le 30/10/2011 18:55, Warren Weckesser a écrit :
Having said that, I think a module specifically for computing
derivatives (with good docs and tests), as being discussed in the ticket #1510 (http://projects.scipy.org/scipy/ticket/1510) would be a nice addition.
Thank to all of you for your responses. I'll follow the discussion on the bug tracker.
Regards,
- -- François Boulogne.
Membre de l'April - Promouvoir et défendre le logiciel libre http://www.april.org -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.11 (GNU/Linux)
iQIcBAEBAgAGBQJOrvbrAAoJEKkn15fZnrRLnRAP/3mB+lbggEegl7Bfvswv1KaO qm8b+TKqoNSNJ42zhq3dg5sRb4RiobAX+nXhS8SOifotnPm4M+BLybEq2bu+TovC 0i83KM86aK7c9fLV5jmN/rhLlFTEZB9Ga722ZYWc5PAg56QSrvjiTPFp4b6cJJLC utOGLCadB7HB6S6m8XzJQ/G66eAMcz2CUCcBvAyOzY+wRLhXLqRqPZhcwrb1QeSf ff8xOQiOAMmcHtnqJfxo6PpLuoItqUmCOXAsfC5yRjdY1AjO82voZ7ZDNuAhhFOl GGJg7z/MEHQRx68gWJ70BxlTIqDhfvc5TI0E7/SFsO4yPd+kYpS5w6Uf6j916WGE X7DdHD6etsqSSpuApx2vXPGgl/ozi4gM+W/H0Ey/8m+KM0N6shkrWEVg8WejUAX0 dfrAR1txo4TioFrx0VwbFtKSsqjyxztbT0nqlO2XSJ5pwGgc6zrLoArJFT1uL25w oMt4pB2UGvLhko8F2LM6Cirr0jfC7c1bjhHkRFpq/8TOvLg+DcRtu9Ag7C1KmskX xpl6s9WtGxFKTPzj5IXNXGboNHaeq454tO5NTd/lRod7r7q0VO1gIldv16AhC7IO UBWfmEIZRXMhKj4ocrRLnyijpUnxZ7ruZBQwMkAJ+SX7pTC5f2/qOFT4ERBjEvyW lL0Qzj8p12lA6DN4Th6t =x1nX -----END PGP SIGNATURE-----
_______________________________________________ SciPy-User mailing list SciPy-User@scipy.org http://mail.scipy.org/mailman/listinfo/scipy-user