多边高斯拟合边界（astropy）

Question

我想做一个多高斯拟合。这里有两个问题，

1. 我可以设置g1，g2和g3，然后将它们全部加起来以获得g？
2. 如何使用边界来设置幅度，平均值和stddev值的限制？或者是否有人知道如何为这三个值设置范围？

我的代码是：

from astropy.modeling.models import Gaussian1D
from astropy.modeling import fitting

np.random.seed(0)
x = x_k
y = y_k
y += np.random.normal(0., 0.2, x.shape)
g1 = Gaussian1D(amplitude=8, mean=-3000, stddev=2000)
g1.bounding_box.amplitude = (5, 10) #I am not sure this format is right because I cannot find an example in astropy
g1.bounding_box.mean = (-4000,-1000)
g2 = Gaussian1D(amplitude=10, mean=0, stddev=2500)
g2.bounding_box.amplitude = (5, 10)
g3 = Gaussian1D(amplitude=8, mean=3000, stddev=2000)
g3.bounding_box.amplitude = (5, 10)
fit_g = fitting.LevMarLSQFitter()
g = fit_g(g1+g2+g3, x, y)

结果是：

g # I input
<CompoundModel9(amplitude_0=5.253073136928387, mean_0=-5457.846704763921, stddev_0=2575.658823137438, amplitude_1=9.341542274082872, mean_1=576.199390319498, stddev_1=4435.00571482351, amplitude_2=4.398784254758611, mean_2=5597.341719866905, stddev_2=1625.151756962933)>

显然它不起作用。怎么了？