[SciPy-User] [SciPy-user] Accessing calculated function values of scipy.interpolated.quad

[Pauli Virtanen]

Mon, 26 Oct 2009 11:51:57 -0400, josef.pktd wrote:

> On Mon, Oct 26, 2009 at 10:40 AM, Richard Pfeifer
> <richard.w.pfeifer at googlemail.com> wrote:
>>
>> Hi Scipy-Users,
>>
>> I've got a question concerning scipy.integrate: I'm interessted in
>>   (1) the Integral F = Int(f, a, b) of the function f(x) in the range
>>   (a,b)
>> AND
>>   (2) the values f(x_i) used for the integration (to plot f(x) in the
>>   same
>> range).
>> scipy.integrate.quad can do the first job very well but I did not find
>> a way to get the values f(x_i). I could let scipy.integrate.quad do
>> it's job and then evaluate f(x) again for the plotting but this would
>> need twice the time...
>>
>> Is there any way to get the [x_i, f(x_i)] scipy.integrate.quad
>> calculated for the integration?
> 
> I think the answer is no.
> Looking at the source of scipy.integrate.quad, the python function
> returns all the return values of the fortran routine. So the actual
> points of evaluation seem to be temp variables inside the fortran
> program.

There's no pre-made solution, but you can cook one yourself:


def func(x):
    return ...long computation...


values = []

def wrapfunc(x):
    y = func(x)
    values.append((x, y))
    return y

ig = integrate.quad(wrapfunc, 0, 1)
values = np.array(values)
values = values[np.argsort(values[:,0])]

-- 
Pauli Virtanen