PRO gfunct, X, A, F, pder
  bx = EXP(A[1] * X)
  F = A[0] * bx + A[2]
 IF N_PARAMS() GE 4 THEN $
    pder = [[bx], [A[0] * X * bx], [replicate(1.0, N_ELEMENTS(X))]]
END
X = FLOAT(INDGEN(10))
Y = [12.0, 11.0, 10.2, 9.4, 8.7, 8.1, 7.5, 6.9, 6.5, 6.1]
 
;Define a vector of weights.
weights = 1.0/Y
 
;Provide an initial guess of the function’s parameters.
A = [10.0,-0.1,2.0]
 
;Compute the parameters.
yfit = CURVEFIT(X, Y, weights, A, SIGMA, FUNCTION_NAME='gfunct')

END