WcsNDMap¶

class gammapy.maps.WcsNDMap(geom, data=None, dtype='float32', meta=None, unit='')[source]¶

Bases: gammapy.maps.WcsMap

HEALPix map with any number of non-spatial dimensions.

This class uses an ND numpy array to store map values. For maps with non-spatial dimensions and variable pixel size it will allocate an array with dimensions commensurate with the largest image plane.

Parameters:	geom : `WcsGeom` WCS geometry object. data : `ndarray` Data array. If none then an empty array will be allocated. dtype : str, optional Data type, default is float32 meta : `dict` Dictionary to store meta data. unit : str or `Unit` The map unit

Attributes Summary

`data`	Data array (`ndarray`)
`geom`	Map geometry (`Geom`)
`meta`	Map meta (`dict`)
`quantity`	Map data times unit (`Quantity`)
`unit`	Map unit (`Unit`)

Methods Summary

`apply_edisp`(self, edisp)	Apply energy dispersion to map.
`coadd`(self, map_in[, weights])	Add the contents of `map_in` to this map.
`convolve`(self, kernel[, use_fft])	Convolve map with a kernel.
`copy`(self, \\kwargs)	Copy map instance and overwrite given attributes, except for geometry.
`create`([map_type, npix, binsz, width, proj, …])	Factory method to create an empty WCS map.
`crop`(self, crop_width)	Crop the spatial dimensions of the map.
`cutout`(self, position, width[, mode])	Create a cutout around a given position.
`downsample`(self, factor[, preserve_counts, axis])	Downsample the spatial dimension by a given factor.
`fill_by_coord`(self, coords[, weights])	Fill pixels at `coords` with given `weights`.
`fill_by_idx`(self, idx[, weights])	Fill pixels at `idx` with given `weights`.
`fill_by_pix`(self, pix[, weights])	Fill pixels at `pix` with given `weights`.
`from_geom`(geom[, meta, data, map_type, …])	Generate an empty map from a `Geom` instance.
`from_hdu`(hdu[, hdu_bands])	Make a WcsNDMap object from a FITS HDU.
`from_hdulist`(hdu_list[, hdu, hdu_bands])	Make a WcsMap object from a FITS HDUList.
`get_by_coord`(self, coords)	Return map values at the given map coordinates.
`get_by_idx`(self, idx)	Return map values at the given pixel indices.
`get_by_pix`(self, pix)	Return map values at the given pixel coordinates.
`get_image_by_coord`(self, coords)	Return spatial map at the given axis coordinates.
`get_image_by_idx`(self, idx)	Return spatial map at the given axis pixel indices.
`get_image_by_pix`(self, pix)	Return spatial map at the given axis pixel coordinates
`get_spectrum`(self[, region, func])	Extract spectrum in a given region.
`interp_by_coord`(self, coords[, interp, …])	Interpolate map values at the given map coordinates.
`interp_by_pix`(self, pix[, interp, fill_value])	Interpolate map values at the given pixel coordinates.
`iter_by_image`(self)	Iterate over image planes of the map.
`make_hdu`(self[, hdu, hdu_bands, sparse, conv])	Make a FITS HDU from this map.
`pad`(self, pad_width[, mode, cval, order])	Pad the spatial dimensions of the map.
`plot`(self[, ax, fig, add_cbar, stretch])	Plot image on matplotlib WCS axes.
`plot_interactive`(self[, rc_params])	Plot map with interactive widgets to explore the non spatial axes.
`read`(filename[, hdu, hdu_bands, map_type])	Read a map from a FITS file.
`reproject`(self, geom[, order, mode])	Reproject this map to a different geometry.
`sample_coord`(self, n_events[, random_state])	Sample position and energy of events.
`set_by_coord`(self, coords, vals)	Set pixels at `coords` with given `vals`.
`set_by_idx`(self, idx, vals)	Set pixels at `idx` with given `vals`.
`set_by_pix`(self, pix, vals)	Set pixels at `pix` with given `vals`.
`slice_by_idx`(self, slices)	Slice sub map from map object.
`smooth`(self, width[, kernel])	Smooth the map.
`sum_over_axes`(self[, keepdims])	To sum map values over all non-spatial axes.
`to_hdulist`(self[, hdu, hdu_bands, sparse, conv])	Convert to `HDUList`.
`upsample`(self, factor[, order, …])	Upsample the spatial dimension by a given factor.
`write`(self, filename[, overwrite])	Write to a FITS file.

Attributes Documentation

data¶: Data array (ndarray)

geom¶: Map geometry (Geom)

meta¶: Map meta (dict)

quantity¶: Map data times unit (Quantity)

unit¶: Map unit (Unit)

Methods Documentation

apply_edisp(self, edisp)[source]¶

Apply energy dispersion to map. Requires energy axis.

Parameters:	edisp : `gammapy.irf.EnergyDispersion` Energy dispersion matrix
Returns:	map : `WcsNDMap` Map with energy dispersion applied.

coadd(self, map_in, weights=None)¶

Add the contents of map_in to this map.

This method can be used to combine maps containing integral quantities (e.g. counts) or differential quantities if the maps have the same binning.

Parameters:	map_in : `Map` Input map. weights: `Map` or `~numpy.ndarray` The weight factors while adding

convolve(self, kernel, use_fft=True, **kwargs)[source]¶

Convolve map with a kernel.

If the kernel is two dimensional, it is applied to all image planes likewise. If the kernel is higher dimensional it must match the map in the number of dimensions and the corresponding kernel is selected for every image plane.

Parameters:	kernel : `PSFKernel` or `numpy.ndarray` Convolution kernel. use_fft : bool Use `scipy.signal.fftconvolve` or `scipy.ndimage.convolve`. kwargs : dict Keyword arguments passed to `scipy.signal.fftconvolve` or `scipy.ndimage.convolve`.
Returns:	map : `WcsNDMap` Convolved map.

copy(self, **kwargs)¶

Copy map instance and overwrite given attributes, except for geometry.

Parameters:	**kwargs : dict Keyword arguments to overwrite in the map constructor.
Returns:	copy : `Map` Copied Map.

classmethod create(map_type='wcs', npix=None, binsz=0.1, width=None, proj='CAR', coordsys='CEL', refpix=None, axes=None, skydir=None, dtype='float32', meta=None, unit='')¶

Factory method to create an empty WCS map.

Parameters:

map_type : {‘wcs’, ‘wcs-sparse’}: Map type. Selects the class that will be used to instantiate the map.
npix : int or tuple or list: Width of the map in pixels. A tuple will be interpreted as parameters for longitude and latitude axes. For maps with non-spatial dimensions, list input can be used to define a different map width in each image plane. This option supersedes width.
width : float or tuple or list: Width of the map in degrees. A tuple will be interpreted as parameters for longitude and latitude axes. For maps with non-spatial dimensions, list input can be used to define a different map width in each image plane.
binsz : float or tuple or list: Map pixel size in degrees. A tuple will be interpreted as parameters for longitude and latitude axes. For maps with non-spatial dimensions, list input can be used to define a different bin size in each image plane.
skydir : tuple or SkyCoord: Sky position of map center. Can be either a SkyCoord object or a tuple of longitude and latitude in deg in the coordinate system of the map.
coordsys : {‘CEL’, ‘GAL’}, optional: Coordinate system, either Galactic (‘GAL’) or Equatorial (‘CEL’).
axes : list: List of non-spatial axes.
proj : string, optional: Any valid WCS projection type. Default is ‘CAR’ (cartesian).
refpix : tuple: Reference pixel of the projection. If None then this will be chosen to be center of the map.
dtype : str, optional: Data type, default is float32
conv : {‘fgst-ccube’,’fgst-template’,’gadf’}, optional: FITS format convention. Default is ‘gadf’.
meta : dict: Dictionary to store meta data.
unit : str or Unit: The unit of the map

Returns:

map : WcsMap: A WCS map object.

crop(self, crop_width)[source]¶

Crop the spatial dimensions of the map.

Parameters:	crop_width : {sequence, array_like, int} Number of pixels cropped from the edges of each axis. Defined analogously to `pad_with` from `numpy.pad`.
Returns:	map : `Map` Cropped map.

cutout(self, position, width, mode='trim')[source]¶

Create a cutout around a given position.

Parameters:	position : `SkyCoord` Center position of the cutout region. width : tuple of `Angle` Angular sizes of the region in (lon, lat) in that specific order. If only one value is passed, a square region is extracted. mode : {‘trim’, ‘partial’, ‘strict’} Mode option for Cutout2D, for details see `Cutout2D`.
Returns:	cutout : `WcsNDMap` Cutout map

downsample(self, factor, preserve_counts=True, axis=None)[source]¶

Downsample the spatial dimension by a given factor.

Parameters:	factor : int Downsampling factor. preserve_counts : bool Preserve the integral over each bin. This should be true if the map is an integral quantity (e.g. counts) and false if the map is a differential quantity (e.g. intensity). axis : str Which axis to downsample. By default spatial axes are downsampled.
Returns:	map : `Map` Downsampled map.

fill_by_coord(self, coords, weights=None)¶

Fill pixels at coords with given weights.

Parameters:	coords : tuple or `MapCoord` Coordinate arrays for each dimension of the map. Tuple should be ordered as (lon, lat, x_0, …, x_n) where x_i are coordinates for non-spatial dimensions of the map. weights : `ndarray` Weights vector. Default is weight of one.

fill_by_idx(self, idx, weights=None)[source]¶

Fill pixels at idx with given weights.

Parameters:	idx : tuple Tuple of pixel index arrays for each dimension of the map. Tuple should be ordered as (I_lon, I_lat, I_0, …, I_n) for WCS maps and (I_hpx, I_0, …, I_n) for HEALPix maps. weights : `ndarray` Weights vector. Default is weight of one.

fill_by_pix(self, pix, weights=None)¶

Fill pixels at pix with given weights.

Parameters:	pix : tuple Tuple of pixel index arrays for each dimension of the map. Tuple should be ordered as (I_lon, I_lat, I_0, …, I_n) for WCS maps and (I_hpx, I_0, …, I_n) for HEALPix maps. Pixel indices can be either float or integer type. Float indices will be rounded to the nearest integer. weights : `ndarray` Weights vector. Default is weight of one.

static from_geom(geom, meta=None, data=None, map_type='auto', unit='', dtype='float32')¶

Generate an empty map from a Geom instance.

Parameters:

geom : Geom: Map geometry.
data : numpy.ndarray: data array
meta : dict: Dictionary to store meta data.
map_type : {‘wcs’, ‘wcs-sparse’, ‘hpx’, ‘hpx-sparse’, ‘auto’}: Map type. Selects the class that will be used to instantiate the map. The map type should be consistent with the geometry. If map_type is ‘auto’ then an appropriate map type will be inferred from type of geom.
unit : str or Unit: Data unit.

Returns:

map_out : Map: Map object

classmethod from_hdu(hdu, hdu_bands=None)[source]¶

Make a WcsNDMap object from a FITS HDU.

Parameters:	hdu : `BinTableHDU` or `ImageHDU` The map FITS HDU. hdu_bands : `BinTableHDU` The BANDS table HDU.

classmethod from_hdulist(hdu_list, hdu=None, hdu_bands=None)¶

Make a WcsMap object from a FITS HDUList.

Parameters:	hdu_list : `HDUList` HDU list containing HDUs for map data and bands. hdu : str Name or index of the HDU with the map data. hdu_bands : str Name or index of the HDU with the BANDS table.
Returns:	wcs_map : `WcsMap` Map object

get_by_coord(self, coords)¶

Return map values at the given map coordinates.

Parameters:	coords : tuple or `MapCoord` Coordinate arrays for each dimension of the map. Tuple should be ordered as (lon, lat, x_0, …, x_n) where x_i are coordinates for non-spatial dimensions of the map.
Returns:	vals : `ndarray` Values of pixels in the map. np.nan used to flag coords outside of map.

get_by_idx(self, idx)[source]¶

Return map values at the given pixel indices.

Parameters:	idx : tuple Tuple of pixel index arrays for each dimension of the map. Tuple should be ordered as (I_lon, I_lat, I_0, …, I_n) for WCS maps and (I_hpx, I_0, …, I_n) for HEALPix maps.
Returns:	vals : `ndarray` Array of pixel values. np.nan used to flag coordinate outside of map

get_by_pix(self, pix)¶

Return map values at the given pixel coordinates.

Parameters:	pix : tuple Tuple of pixel index arrays for each dimension of the map. Tuple should be ordered as (I_lon, I_lat, I_0, …, I_n) for WCS maps and (I_hpx, I_0, …, I_n) for HEALPix maps. Pixel indices can be either float or integer type.
Returns:	vals : `ndarray` Array of pixel values. np.nan used to flag coordinates outside of map

get_image_by_coord(self, coords)¶

Return spatial map at the given axis coordinates.

Parameters:	coords : tuple or dict Tuple should be ordered as (x_0, …, x_n) where x_i are coordinates for non-spatial dimensions of the map. Dict should specify the axis names of the non-spatial axes such as {‘axes0’: x_0, …, ‘axesn’: x_n}.
Returns:	map_out : `Map` Map with spatial dimensions only.

See also

get_image_by_coord, get_image_by_idx

get_spectrum(self, region=None, func=<function nansum at 0x10b3a9840>)[source]¶

Extract spectrum in a given region.

The spectrum can be computed by summing (or, more generally, applying func) along the spatial axes in each energy bin. This occurs only inside the region, which by default is assumed to be the whole spatial extension of the map.

Parameters:	region: `~regions.Region` Region (pixel or sky regions accepted). func : numpy.ufunc Function to reduce the data.
Returns:	spectrum : `CountsSpectrum` Spectrum in the given region.

interp_by_coord(self, coords, interp=None, fill_value=None)[source]¶

Interpolate map values at the given map coordinates.

Parameters:

coords : tuple or MapCoord: Coordinate arrays for each dimension of the map. Tuple should be ordered as (lon, lat, x_0, …, x_n) where x_i are coordinates for non-spatial dimensions of the map.
interp : {None, ‘nearest’, ‘linear’, ‘cubic’, 0, 1, 2, 3}: Method to interpolate data values. By default no interpolation is performed and the return value will be the amplitude of the pixel encompassing the given coordinate. Integer values can be used in lieu of strings to choose the interpolation method of the given order (0=’nearest’, 1=’linear’, 2=’quadratic’, 3=’cubic’). Note that only ‘nearest’ and ‘linear’ methods are supported for all map types.
fill_value : None or float value: The value to use for points outside of the interpolation domain. If None, values outside the domain are extrapolated.

Returns:

vals : ndarray: Interpolated pixel values.

interp_by_pix(self, pix, interp=None, fill_value=None)[source]¶: Interpolate map values at the given pixel coordinates.

iter_by_image(self)¶

Iterate over image planes of the map.

This is a generator yielding (data, idx) tuples, where data is a numpy.ndarray view of the image plane data, and idx is a tuple of int, the index of the image plane.

The image plane index is in data order, so that the data array can be indexed directly. See Iterating by image for further information.

make_hdu(self, hdu='SKYMAP', hdu_bands=None, sparse=False, conv=None)¶

Make a FITS HDU from this map.

Parameters:	hdu : str The HDU extension name. hdu_bands : str The HDU extension name for BANDS table. sparse : bool Set INDXSCHM to SPARSE and sparsify the map by only writing pixels with non-zero amplitude.
Returns:	hdu : `BinTableHDU` or `ImageHDU` HDU containing the map data.

pad(self, pad_width, mode='constant', cval=0, order=1)[source]¶

Pad the spatial dimensions of the map.

Parameters:

pad_width : {sequence, array_like, int}: Number of pixels padded to the edges of each axis.
mode : {‘edge’, ‘constant’, ‘interp’}: Padding mode. ‘edge’ pads with the closest edge value. ‘constant’ pads with a constant value. ‘interp’ pads with an extrapolated value.
cval : float: Padding value when mode=’consant’.
order : int: Order of interpolation when mode=’constant’ (0 = nearest-neighbor, 1 = linear, 2 = quadratic, 3 = cubic).

Returns:

map : Map: Padded map.

plot(self, ax=None, fig=None, add_cbar=False, stretch='linear', **kwargs)[source]¶

Plot image on matplotlib WCS axes.

Parameters:	ax : `WCSAxes`, optional WCS axis object to plot on. fig : `Figure` Figure object. add_cbar : bool Add color bar? stretch : str Passed to `astropy.visualization.simple_norm`. **kwargs : dict Keyword arguments passed to `imshow`.
Returns:	fig : `Figure` Figure object. ax : `WCSAxes` WCS axis object cbar : `Colorbar` or None Colorbar object.

plot_interactive(self, rc_params=None, **kwargs)¶

Plot map with interactive widgets to explore the non spatial axes.

Parameters:	rc_params : dict Passed to `matplotlib.rc_context(rc=rc_params)` to style the plot. **kwargs : dict Keyword arguments passed to `WcsNDMap.plot`.

Examples

You can try this out e.g. using a Fermi-LAT diffuse model cube with an energy axis:

from gammapy.maps import Map

m = Map.read("$GAMMAPY_DATA/fermi_3fhl/gll_iem_v06_cutout.fits")
m.plot_interactive(add_cbar=True, stretch="sqrt")

If you would like to adjust the figure size you can use the rc_params argument:

rc_params = {'figure.figsize': (12, 6), 'font.size': 12}
m.plot_interactive(rc_params=rc_params)

static read(filename, hdu=None, hdu_bands=None, map_type='auto')¶

Read a map from a FITS file.

Parameters:

filename : str or Path: Name of the FITS file.
hdu : str: Name or index of the HDU with the map data.
hdu_bands : str: Name or index of the HDU with the BANDS table. If not defined this will be inferred from the FITS header of the map HDU.
map_type : {‘wcs’, ‘wcs-sparse’, ‘hpx’, ‘hpx-sparse’, ‘auto’}: Map type. Selects the class that will be used to instantiate the map. The map type should be consistent with the format of the input file. If map_type is ‘auto’ then an appropriate map type will be inferred from the input file.

Returns:

map_out : Map: Map object

reproject(self, geom, order=1, mode='interp')¶

Reproject this map to a different geometry.

Only spatial axes are reprojected, if you would like to reproject non-spatial axes consider using Map.interp_by_coord() instead.

Parameters:	geom : `Geom` Geometry of projection. mode : {‘interp’, ‘exact’} Method for reprojection. ‘interp’ method interpolates at pixel centers. ‘exact’ method integrates over intersection of pixels. order : int or str Order of interpolating polynomial (0 = nearest-neighbor, 1 = linear, 2 = quadratic, 3 = cubic).
Returns:	map : `Map` Reprojected map.

sample_coord(self, n_events, random_state=0)[source]¶

Sample position and energy of events.

Parameters:	n_events : int Number of events to sample. random_state : {int, ‘random-seed’, ‘global-rng’, `RandomState`} Defines random number generator initialisation. Passed to `get_random_state`.
Returns:	coords : `MapCoord` object. Sequence of coordinates and energies of the sampled events.

set_by_coord(self, coords, vals)¶

Set pixels at coords with given vals.

Parameters:	coords : tuple or `MapCoord` Coordinate arrays for each dimension of the map. Tuple should be ordered as (lon, lat, x_0, …, x_n) where x_i are coordinates for non-spatial dimensions of the map. vals : `ndarray` Values vector.

set_by_idx(self, idx, vals)[source]¶

Set pixels at idx with given vals.

Parameters:	idx : tuple Tuple of pixel index arrays for each dimension of the map. Tuple should be ordered as (I_lon, I_lat, I_0, …, I_n) for WCS maps and (I_hpx, I_0, …, I_n) for HEALPix maps. vals : `ndarray` Values vector.

set_by_pix(self, pix, vals)¶

Set pixels at pix with given vals.

Parameters:	pix : tuple Tuple of pixel index arrays for each dimension of the map. Tuple should be ordered as (I_lon, I_lat, I_0, …, I_n) for WCS maps and (I_hpx, I_0, …, I_n) for HEALPix maps. Pixel indices can be either float or integer type. Float indices will be rounded to the nearest integer. vals : `ndarray` Values vector.

slice_by_idx(self, slices)¶

Slice sub map from map object.

For usage examples, see Indexing and Slicing.

Parameters:	slices : dict Dict of axes names and integers or `slice` object pairs. Contains one element for each non-spatial dimension. For integer indexing the corresponding axes is dropped from the map. Axes not specified in the dict are kept unchanged.
Returns:	map_out : `Map` Sliced map object.

smooth(self, width, kernel='gauss', **kwargs)[source]¶

Smooth the map.

Iterates over 2D image planes, processing one at a time.

Parameters:

width : Quantity, str or float: Smoothing width given as quantity or float. If a float is given it interpreted as smoothing width in pixels. If an (angular) quantity is given it converted to pixels using geom.wcs.wcs.cdelt. It corresponds to the standard deviation in case of a Gaussian kernel, the radius in case of a disk kernel, and the side length in case of a box kernel.
kernel : {‘gauss’, ‘disk’, ‘box’}: Kernel shape
kwargs : dict: Keyword arguments passed to uniform_filter (‘box’), gaussian_filter (‘gauss’) or convolve (‘disk’).

Returns:

image : WcsNDMap: Smoothed image (a copy, the original object is unchanged).

sum_over_axes(self, keepdims=False)[source]¶

To sum map values over all non-spatial axes.

Parameters:	keepdims : bool, optional If this is set to true, the axes which are summed over are left in the map with a single bin
Returns:	map_out : WcsNDMap Map with non-spatial axes summed over

to_hdulist(self, hdu=None, hdu_bands=None, sparse=False, conv='gadf')¶

Convert to HDUList.

Parameters:	hdu : str Name or index of the HDU with the map data. hdu_bands : str Name or index of the HDU with the BANDS table. sparse : bool Sparsify the map by only writing pixels with non-zero amplitude. conv : {‘gadf’, ‘fgst-ccube’,’fgst-template’} FITS format convention.
Returns:	hdu_list : `HDUList`

upsample(self, factor, order=0, preserve_counts=True, axis=None)[source]¶

Upsample the spatial dimension by a given factor.

Parameters:	factor : int Upsampling factor. order : int Order of the interpolation used for upsampling. preserve_counts : bool Preserve the integral over each bin. This should be true if the map is an integral quantity (e.g. counts) and false if the map is a differential quantity (e.g. intensity). axis : str Which axis to upsample. By default spatial axes are upsampled.
Returns:	map : `Map` Upsampled map.

write(self, filename, overwrite=False, **kwargs)¶

Write to a FITS file.

Parameters:

filename : str: Output file name.
overwrite : bool: Overwrite existing file?
hdu : str: Set the name of the image extension. By default this will be set to SKYMAP (for BINTABLE HDU) or PRIMARY (for IMAGE HDU).
hdu_bands : str: Set the name of the bands table extension. By default this will be set to BANDS.
conv : str: FITS format convention. By default files will be written to the gamma-astro-data-formats (GADF) format. This option can be used to write files that are compliant with format conventions required by specific software (e.g. the Fermi Science Tools). Supported conventions are ‘gadf’, ‘fgst-ccube’, ‘fgst-ltcube’, ‘fgst-bexpcube’, ‘fgst-template’, ‘fgst-srcmap’, ‘fgst-srcmap-sparse’, ‘galprop’, and ‘galprop2’.
sparse : bool: Sparsify the map by dropping pixels with zero amplitude. This option is only compatible with the ‘gadf’ format.