box

The LT box contains an assortment of plotting, histogramming and fitting routines used for data analysis. The plotting routines have been imported from the plotting module and the fitting routines have been imported from the The LT_Fit Package. For detailed information about these functions check the plotting and the the The LT_Fit Package documentation.

Normally all you should need to do is import the box and you have a box of tools.

Example:

>>> import LT.box as B

Imported functions:

From LT.plotting:
From LT_Fit:

Box Content:

LT.box.get_file(file, **kwargs)

Assume that B is the name of LT.box.

Open and read the file:

>>> md = B.get_file( 'file.data' )
LT.box.get_data(D, var)

Assume that B is the name of LT.box.

Get all the values of variable ‘my_var’ as a numpy.array():

>>> mv = B.get_data(md, 'my_var')

This is an old method, better use:

>>> mv = md['my_var']
LT.box.get_spectrum(file, calibration=None)

Read an MCA spectrum and convert it to a histogram.

Example:

>>> h = B.get_spectrum('co60.spe')  # read the spectrum
>>> h.plot()                        # plot the spectrum
class LT.box.histo(values=None, range=None, bins=None, histogram=None, bin_error=None, bin_center=None, bin_content=None, file=None, window=None, title='my histogram', xlabel='x-bin', ylabel='content', calc_w2=False, **kwargs)

Define a histogram based on the np.histogram class.

The various ways of defining one are:

  • If a is a 1D ( numpy.array()) containing the data to be histogrammed

    >>> h = histo( a )

  • If his is the output of the numpy.histogram() function

    >>> h = histo(histogram = his)

  • If bc is a 1D array with bin center values, and bcont contains bin content values then:

    >>> h = histo(bin_center = bc, bin_content = bcont)

  • A filename for a stored histogram is given

    >>> h = histo(filename), where filename contains the pdatafile

    Usually the result of a histo.save operation

Important keywords:

Keyword

Meaning

values

Array of values to be histogrammed (numpy.array())

range

Lower and upper limits of binning ( e.g. range = (10.,20.) )

bins

Number of bins

histogram

Result of numpy.histogram() function

bin_error

Array of errors for each bin content (numpy.array())

bin_center

Array of bin-center values (numpy.array())

bin_content

Array of bin-content values (numpy.array())

file

Load data from file

window

Set a window (a zoom window)

title

Set the title

xlabel

Set the x-label

ylabel

Set the y-label

calc_w2

If True calccultae the sum of the square of the weights for each bin to calculate the final bin error

weights

Array of weight to increment the histogram with

Additional keyword arguments are passed to the numpy.histogram() function

apply_calibration(cal)

apply x-axis calibration, new axis values are cal(xaxis)

bkg(x)

Return the value of the fitted background at x

Parameters:
xfloat
Returns:
b_val(x)float

fitted backgound at x.

clear()

Set the content and errors to 0.

clear_window()

Reset (Clear) the defined window

copy()

Create a copy of the histogram:

>>>hc = h.copy()

Only the histogram values are copied, no lables and titles etc.

fill(y, add=False, **kwargs)

Fill the histogram with the values stored in the numpy.array() y.

Keyword

Meaning

add

if True add the results to the existing content

Additional keyword arguments are passed to the numpy.histogram() function

find_bin(x)

Find the bin that would contain the value x

fit(xmin=None, xmax=None, init=True, ignore_zeros=True, **kwargs)

Fit a gaussian on a quadratic background. You can also just fit a background or just a gaussian. All this is controlled by which parameters you want to fit. Another important part of non-linear fitting is that you need to provide reasonable guesses for the fit parameters. The parameters in histo are not just numbers but objects with their own properties and functions (see Parameter ). The full fit function is as follows:

$ f(x) = b_0 + b_1x + b_2x^2 + A exp(-(x - \mu)^2/\sigma^2)$

The (LT.box.histo) parameters are:

Parameter

Histo Class Member

$b_o $

b0

$b_1 $

b1

$b_2 $

b2

$A $

A

$\mu $

mean

$\sigma $

sigma

Which parameters are fitted is defined in set_fit_list()

Keyword arguments are:

Keyword

Meaning

xmin

lower fit limit

xmax

upper fit limit

init

True/False (default = True) estimate initial fit parameters automatically

ignore_zeros

True/False (default = True) ignore channels with bin content zero

kwargs

additional keywords are passed to gen_fit (use only if you know what you are doing!)
fit_func(x)

The function fitted to the histogram data

fit_view(init=True)

Fit histogram using the current display limits as fit range. This is only useful if the histogram has been plotted.

gauss(x)

Return the value of the gauss function using the fitted parameters at x

Parameters:
xfloat
Returns:
gauss(x)float

fitted gaussian at x.

init_gauss(xmin=None, xmax=None, ignore_zeros=True)

Calculate the initial parameter guess for a gaussian. These parameters can them be used in the call to fit

init_parameters()

Reset fit parameters to their default values

Returns:
None.
load(file='histo.data', clear_window=False)

read the histogram data from pdatafile

If the file does not result from a save function make sure that all the necessary data are present.

plot(filled='True', ymin=0.0, axes=None, ignore_zeros=False, **kwargs)

Plot the histogram content:

Keyword

Meaning

filled

if True draw a filled histogram

ymin

lower limit where fill starts (horizontal line)

ignore_zeros

do not plot channels with bin content 0 (default = False)
plot_exp(ignore_zeros=False, axes=None, **kwargs)

Plot histogram content and errors like experimental data.

Keyword

Meaning

ignore_zeros

do not plot channels with bin content 0 (default = False)
plot_fit(color='r', axes=None, **kwargs)

Plot the fitted curve

Keyword

Meaning

color

color of the fitted line
plot_guess(xmin=None, xmax=None)

Plot the fit function with the guessed parameters

Keyword arguments are:

Keyword

Meaning

xmin

lower fit limit

xmax

upper fit limit
rebin(n, scale=False, use_mean=False, replace=False)

rebin the histogram by a factor n:

>>>hc = h.rebin(2)

Keyword

Meaning

scale
True: the original bin number is not a multiple of n

and the last bin content will be scaled accordingly

use_mean
True: the new bin content is the mean of the bin_content

of n bins

replace
True: replace the current histogram with the

rebinned version
save(filename='histo.data')

Save the histogram in pdatafile format

set_fit_list(fit=['A', 'mean', 'sigma'])

Define which parameters are to be fitted.

The default list is

fit = [ 'A', 'mean', 'sigma']

to use all parameters:

h.set_fit_list( fit = [ 'A', 'mean', 'sigma', 'b0', 'b1', 'b2'])
set_window(xmin=None, xmax=None)

Define a window into the histogram. This is similar to a zoom or a region of interest (ROI)

Keyword

Meaning

xmin

lower limit for plotting or fitting

xmax

upper limit
set_window_view()

Like set_windows but uses the current display limits. This is only useful if the histogram has been plotted.

show_fit_list()

Show the current fit list

Returns:
None.
sum(xmin=None, xmax=None)

Return the sum of all bins. If the limits are given, calculate the sum of all bins between the bins that contain the values xmin and xmax. If a window is set only the sum within the window is returned when no limits are specified.

Example:

>>> s0 = h.sum() # add all bins
>>> s1 = h.sum(0.5, 1.1) # add the bins between 0.5 and 1.1
>>> s2 = h.sum(xmin = 0.5, xmax = 1.1) # add the bins between 0.5 and 1.1

Keyword

Meaning

xmin

lower limit of sum of bin content

xmax

upper limit of sum

The errors are also calculated.

class LT.box.histo2d(x_values=None, y_values=None, range=None, bins=None, histogram=None, bin_error=None, bin_content=None, x_bin_center=None, y_bin_center=None, file=None, title='my histogram', xlabel='x-bin', ylabel='y-bin', zlabel='content', bad_color='w', colorbar=True, logz=False, calc_w2=False, **kwargs)

Define a 2d histogram based on the np.histogram2d class.

The various ways of defining one are:

  • If xv is a 1D ( numpy.array()) containing the x-value data to be histogrammed and yx is a 1D array containing the y-value data :

    >>> h2 = histo2d( xv, yv )

  • If his2 is the output of the numpy.histogram2d() function

    >>> h2 = histo2d(histogram = his2)

  • If xbc is an array with x-bin center values, ybc is an array with y-bin center values and bcont contains bin content values then use:

    >>> h2 = histo2d(x_bin_center = xbc, y_bin_center = ybc, bin_content = bcont)

  • A filename for a stored histogram is given

    >>> h2 = histo2d(filename), where filename contains the pdatafile

    Usually the result of a histo2d.save operation

Important keywords:

Keyword

Meaning

xvalues

Array of x-values to be histogrammed (1d-numpy.array())

yvalues

Array of y-values to be histogrammed (1d-numpy.array())

range

Lower and upper limits of binning ( e.g. range = (10.,20.) )

bins

Number of bins, or [binsx, binsy]

histogram

Result of numpy.histogram2d() function

x_bin_center

1d-Array of x - bin-center values (numpy.array())

y_bin_center

1d-Array of y - bin-center values (numpy.array())

bin_content

2d-Array of bin-content values (numpy.array())

bin_error

2d-Array of errors for each bin content (numpy.array())

file

Load data from file

title

Set the title

xlabel

Set the x-label

ylabel

Set the y-label

zlabel

Set the z-label

calc_w2

If True calccultae the sum of the square of the weights for each bin to calculate the final bin error

weights

Array of weight to increment the histogram with

colorbar

if True, plot a colorbar

bad_color

Set the color for plot for bins below zmin (default: w)

logz

if True plot content on log scale

Additional keyword arguments are passed to the numpy.histogram2d() function

apply_calibration(cal_x, cal_y)

apply x and y-axis calibration, new axis values are cal(xaxis) cal(yaxis)

Keyword

Meaning

cal_x

x-axis calibration function

cal_y

y-axis calibration function
Returns:
None., calculates new bin vakues
clear()

Set the content and errors to 0.

draw_polygon(p, **kwargs)

draw a polygon used as cut

Keyword

Meaning

p

array of coordinate pairs determining the corners of the polygon
Returns:
None.
fill(x, y, add=False, **kwargs)

Fill the histogram with the values stored in the numpy.array() y.

Keyword

Meaning

add

if True add the results to the existing content

Additional keyword arguments are passed to the numpy.histogram() function

find_bin(x, y)

Find the bin value pair for that would contain the value pair x, y

plot(axes=None, graph='patch', clevel=10, colormap=<matplotlib.colors.LinearSegmentedColormap object>, logz=False, skip_x_label=False, skip_ylabel=False, skip_title=False, **kwargs)

Plot the 2d histogram content:

Keyword

Meaning

graph

type of plot: patch, contour, surface, lego

clevel

number of contour levels (default 10)

colormap

colormap to be used (default CMRmap)

logz

if True use a logarithmic scale for content (patch and contour only)

skip_x_label

if True do not plot x-label

skip_y_label

if True do not plot y-label

skip_title

if True do not plot title

kwargs

additional kwargs are possed to the plotting routines
poly_cut(p)

Setup a polygon cut :

Keyword

Meaning

p

array of coordinate pairs determining the corners of the polygon
example for a triangle cut :

p = np.array( [[1,2], [2,4], [0.5,3] ])

Returns:
Masked array to be applied to the bin_content and bin_error
project_x(range=None, bins=None)

project a range of y-bins onto the x-axis

Keyword

Meaning

range

the range in y included in the projection

bins

an array of bins or a slice selecting the y-bins to be included in the projection

both None

Project all the y-bins
Returns:
1d - histogram
project_y(range=None, bins=None)

project a range of x-bins onto the y-axis

Keyword

Meaning

range

the range in x included in the projection

bins

an array of bins or a slice selecting the x-bins to be included in the projection

both None

Project all the x-bins
Returns:
1d - histogram
rect_cut(x1, x2, y1, y2)

Setup a rectangle cut

Keyword

Meaning

x1

lower x limit

x2

upper x limit

y1

lower y limit

y2

upper y limit
Returns:
selmasked array for the bin_content
save(filename='histo2d.data', ignore_zeros=True)

Save the histogram in pdatafile format

Keyword

Meaning

filename:

filename to be used

ignore_zeros

if True, write only bins with non-zero content (default True)
set_nans(value=0.0, err_value=1.0)

replace nans by specified values

Keyword

Meaning

value

replaces the nan in bin_content (default 0)

err_value

replaces the corresponding bin_error (default 1)
Returns:
None.
sum(rect_cut=None, poly_cut=None, draw=False)

Return the sum of all bins. If the limits are of a rectangle are given, calculate the sum of all bins inside it or inside a polygon poly

Example:

>>> s0 = h.sum() # add all bins
>>> s1 = h.sum(rect = [x1,x2, y1, y2])
>>> s2 = h.sum(poly = ([x1, x2, ...], [y1, y2, ...])) # add the bins inside the polygon

Keyword

Meaning

xmin

lower limit of sum of bin content

xmax

upper limit of sum

The errors are also calculated.

Histogram Operations

The following operations are defined for Histograms (1-D and 2-D) (histo ):

In [1]: hr = h1 + h2        #  addition
In [2]: hr = h1 - h2        # subtraction
In [3]: hr = 2.*h1          # mutiplication by a number
In [4]: hr = 5.5*h1 - 2.*h2 # combination of the above
Note: When adding histograms they need to be of the same type (the

same number of bins, same bin-center values). Labels and titles are not copied.