添加或减去 2 条插值曲线的最佳方法是什么？

Question

在 Python 中添加或减去 2 条插值曲线的最佳方法是什么。一个例子是在国债利率上增加信用利差。两条曲线没有相同的项点。我可以通过获得特定的男高音来重建曲线，但我希望有更好的方法。

import scipy as sc
ATenor = [0.25,0.5,1,5,10,20]
ARate = [0.02,0.022,0.025,0.03,0.035,0.039]
ACurve = sc.interpolate.interp1d(ATenor,ARate)
BTenor = [0.25,1,4,5,7,10,15,20]
BRate = [0.025,0.28,0.032,0.036,0.038,0.042,0.04,0.038]
BCurve = sc.interpolate.interp1d(BTenor,BRate)
CCurve = ACurve + BCurve  # <-- This does not work but to get the idea across

如果它尝试添加 2 条插值曲线，我会收到此错误：

+ 不支持的操作数类型：'interp1d' 和 'interp1d'

Answer 1

如果要创建新的 interp1d 对象，则可以合并 tenor 数组 (x-axis) 并重新计算速率值 (y-axis)。例如，这段代码将执行：

ABTenor = sorted(set(ATenor + BTenor))              # Merge points on the x-axis.
ABCurve = [ACurve(x) + BCurve(x) for x in ABTenor]  # Compute y values.
ABCurve = scipy.interpolate.interp1d(ABTenor, ABCurve)

或者，您可以创建惰性求值函数：

def add_curves(A, B):
    def compute(x):
        return A(x) + B(x)
    return compute

ABCurve = add_curves(ACurve, BCurve)
# ABCurve(10.0) will call ACurve(10.0) and BCurve(10.0) and sum the results.

添加或减去 2 条插值曲线的最佳方法是什么？

What's the best way to add or subtract 2 interpolated curves?

python

interpolation

scipy