healpy.pixelfunc.get_interp_weights

healpy.pixelfunc.get_interp_weights(nside, theta, phi=None, nest=False, lonlat=False)

Return the 4 closest pixels on the two rings above and below the location and corresponding weights. Weights are provided for bilinear interpolation along latitude and longitude

Parameters

nsideint

the healpix nside

theta, phifloat, scalar or array-like

if phi is not given, theta is interpreted as pixel number, otherwise theta[rad],phi[rad] are angular coordinates

nestbool

if True, NESTED ordering, otherwise RING ordering.

lonlatbool

If True, input angles are assumed to be longitude and latitude in degree, otherwise, they are co-latitude and longitude in radians.

Returns

restuple of length 2

contains pixel numbers in res[0] and weights in res[1]. Usual numpy broadcasting rules apply.

See also

get_interp_val, get_all_neighbours

Examples

Note that some of the test inputs below that are on pixel boundaries such as theta=pi/2, phi=pi/2, have a tiny value of 1e-15 added to them to make them reproducible on i386 machines using x87 floating point instruction set (see https://github.com/healpy/healpy/issues/528).

>>> import healpy as hp
>>> pix, weights = hp.get_interp_weights(1, 0)
>>> print(pix)
[0 1 4 5]
>>> weights
array([ 1.,  0.,  0.,  0.])

>>> pix, weights = hp.get_interp_weights(1, 0, 0)
>>> print(pix)
[1 2 3 0]
>>> weights
array([ 0.25,  0.25,  0.25,  0.25])

>>> pix, weights = hp.get_interp_weights(1, 0, 90, lonlat=True)
>>> print(pix)
[1 2 3 0]
>>> weights
array([ 0.25,  0.25,  0.25,  0.25])

>>> pix, weights = hp.get_interp_weights(1, [0, np.pi/2 + 1e-15], 0)
>>> print(pix)
[[ 1  4]
 [ 2  5]
 [ 3 11]
 [ 0  8]]
>>> np.testing.assert_allclose(
...     weights,
...     np.array([[ 0.25,  1.  ],
...               [ 0.25,  0.  ],
...               [ 0.25,  0.  ],
...               [ 0.25,  0.  ]]), rtol=0, atol=1e-14)