| | |
- SIGN(a, b)
- bfgs(x, f, df, costTol=1e-10, gradTol=0.0001, maxIters=200, verbose=0, stepsize0=1, resetInterval=None)
Perform Broyden-Fletcher-Goldfarb-Shanno variant of
Davidson-Fletcher-Powell minimization.
Return value is the tuple (x,J(x),numIters)
- brent(ax, bx, cx, f, tol=1e-08, ITMAX=100, ZEPS=1e-10)
Find minimum bracketed by ax,bx,cx
return (xmin,f(xmin))
- conmin(x, f, df, costTol=1e-10, gradTol=0.0001, maxIters=200, stepsize=1, verbose=0)
conjugate gradient minimization of function f with gradient df, starting at
position x.
stepsize specifies initial stepsize to be taken
Return value is the tuple (x,J(x),numIters)
FIX: this routine makes unnecessary function calls.
- dot(...)
- linemin(func, p, xi, stepsize)
determine the minimum of func starting at point p, in the direction xi.
return (pf,pfmin,xif)
minimum position, fmin, and the actual displacement vector used
- mnbrak(ax, bx, func)
- given function func, and given distinct initial points ax and bx, this
routine searches in the downhill direction *defined by the function as
evaluated at the initial points) and returns new points ax, bx, cx which
bracket a minimum of the function. Also returned are the function values
at the three points fa, fb and fc.
- norm(...)
- outerProd(x, y)
return the outer product matrix result of multiplying the vectors x and y
|