ImageProfile

class gammapy.image.ImageProfile(table)

Bases: object

Image profile class.

The image profile data is stored in Table object, with the following columns:

• x_ref Coordinate bin center (required).
• x_min Coordinate bin minimum (optional).
• x_max Coordinate bin maximum (optional).
• profile Image profile data (required).
• profile_err Image profile data error (optional).

Parameters:

table : Table

Table instance with the columns specified as above.

Attributes Summary

profile	Image profile quantity.
profile_err	Image profile error quantity.
x_max	Max.
x_min	Min.
x_ref	Reference x coordinates.

Methods Summary

normalize([mode])	Normalize profile to peak value or integral.
peek([figsize])	Show image profile and error.
plot([ax])	Plot image profile.
plot_err([ax])	Plot image profile error as band.
smooth([kernel, radius])	Smooth profile with error propagation.

Attributes Documentation

profile

Image profile quantity.

profile_err

Image profile error quantity.

x_max

Max. x coordinates.

x_min

Min. x coordinates.

x_ref

Reference x coordinates.

Methods Documentation

normalize(mode='peak')

Normalize profile to peak value or integral.

Parameters:

mode : [‘integral’, ‘peak’]

Normalize image profile so that it integrates to unity (‘integral’) or the maximum value corresponds to one (‘peak’).

Returns:

profile : ImageProfile

Normalized image profile.

peek(figsize=(8, 4.5), **kwargs)

Show image profile and error.

Parameters:

**kwargs : dict

Keyword arguments passed to ImageProfile.plot_profile()

Returns:

ax : Axes

Axes object

plot(ax=None, **kwargs)

Plot image profile.

Parameters:

ax : Axes

Axes object

**kwargs : dict

Keyword arguments passed to plot

Returns:

ax : Axes

Axes object

plot_err(ax=None, **kwargs)

Plot image profile error as band.

Parameters:

ax : Axes

Axes object

**kwargs : dict

Keyword arguments passed to plt.fill_between()

Returns:

ax : Axes

Axes object

smooth(kernel='box', radius='0.1 deg', **kwargs)

Smooth profile with error propagation.

Smoothing is described by a convolution:

\[x_j = \sum_i x_{(j - i)} h_i\]

Where \(h_i\) are the coefficients of the convolution kernel.

The corresponding error on \(x_j\) is then estimated using Gaussian error propagation, neglecting correlations between the individual \(x_{(j - i)}\):

\[\Delta x_j = \sqrt{\sum_i \Delta x^{2}_{(j - i)} h^{2}_i}\]

Parameters:

kernel : {‘gauss’, ‘box’}

Kernel shape

radius : Quantity, str or float

Smoothing width given as quantity or float. If a float is given it is interpreted as smoothing width in pixels. If an (angular) quantity is given it is converted to pixels using xref[1] - x_ref[0].

kwargs : dict

Keyword arguments passed to uniform_filter (‘box’) and gaussian_filter (‘gauss’).

Returns:

profile : ImageProfile

Smoothed image profile.