I want to find a cubic spline function approximation given data ${x_i,y_i}$ which is equivalent to scipy.interpolate.interp1d(x,y,kind=‘cubic’). Scipy gives a numpy-type function and this will disconnect graph. I tried to find such a function but I think torch.nn.functional.intepolate is different from what I want. Is it available in pytorch?
did you ever find a solution to this?
Love you so much!
That’s exactly what I want!