Utilities

poly(p, x)

Polynomial function for curve fitting. Uses Horner's method via evalpoly.

Arguments

• p: Coefficients [c₀, c₁, c₂, ...] for c₀ + c₁x + c₂x² + ...
• x: Independent variable

Example

poly([1.0, 2.0, 3.0], [0.0, 1.0, 2.0])  # 1 + 2x + 3x²
# returns [1.0, 6.0, 17.0]

Combine with other models for simultaneous baseline fitting:

model = combine(lorentzian, 3, poly, 2)

combine(f1, n1, f2, n2)

Combine two model functions into one by splitting the parameter vector.

Arguments

• f1: First model function with signature f1(p, x)
• n1: Number of parameters for f1
• f2: Second model function with signature f2(p, x)
• n2: Number of parameters for f2

Returns a new function (p, x) -> f1(p[1:n1], x) .+ f2(p[n1+1:n1+n2], x).

Example

# Lorentzian (3 params) + linear baseline (2 params)
model = combine(lorentzian, 3, poly, 2)

p0 = [1.0, 0.0, 1.0, 0.1, 0.01]  # [A, x0, Γ, c0, c1]
prob = NonlinearCurveFitProblem(model, p0, x, y)
sol = solve(prob)