def test_guess_from_peak2d():
"""Regression test for guess_from_peak2d function (see GH #627)."""
x = np.linspace(-5, 5)
y = np.linspace(-5, 5)
amplitude = 0.8
centerx = 1.7
sigmax = 0.3
centery = 1.3
sigmay = 0.2
z = lineshapes.gaussian2d(x,
y,
amplitude=amplitude,
centerx=centerx,
sigmax=sigmax,
centery=centery,
sigmay=sigmay)
model = models.Gaussian2dModel()
guess_increasing_x = model.guess(z, x=x, y=y)
guess_decreasing_x = model.guess(z[::-1], x=x[::-1], y=y[::-1])
assert guess_increasing_x == guess_decreasing_x
for param, value in zip(['centerx', 'centery'], [centerx, centery]):
assert np.abs((guess_increasing_x[param].value - value) / value) < 0.5
from lmfit.lineshapes import gaussian2d, lorentzian
###############################################################################
# Two-dimensional Gaussian
# ------------------------
# We start by considering a simple two-dimensional gaussian function, which
# depends on coordinates `(x, y)`. The most general case of experimental
# data will be irregularly sampled and noisy. Let's simulate some:
npoints = 10000
np.random.seed(2021)
x = np.random.rand(npoints) * 10 - 4
y = np.random.rand(npoints) * 5 - 3
z = gaussian2d(x,
y,
amplitude=30,
centerx=2,
centery=-.5,
sigmax=.6,
sigmay=.8)
z += 2 * (np.random.rand(*z.shape) - .5)
error = np.sqrt(z + 1)
###############################################################################
# To plot this, we can interpolate the data onto a grid.
X, Y = np.meshgrid(np.linspace(x.min(), x.max(), 100),
np.linspace(y.min(), y.max(), 100))
Z = griddata((x, y), z, (X, Y), method='linear', fill_value=0)
fig, ax = plt.subplots()
art = ax.pcolor(X, Y, Z, shading='auto')
plt.colorbar(art, ax=ax, label='z')
ax.set_xlabel('x')
