fitspy.baseline module

Class dedicated to spectrum baseline manipulation

class fitspy.baseline.BaseLine

Bases: object

Class dedicated to spectrum baseline manipulation

points

List of the (x,y) baseline points coordinates

Type:: list of 2 lists

Mode used to determine the baseline, among ‘Semi-Auto’ (semi-automatic baseline determination), ‘Linear’ (piecewise linear decomposition based on users points definition) and ‘Polynomial’. Default mode is ‘Linear’

Type:: str

coef

Smoothing coefficient used in the ‘Semi-Auto’ mode. The larger coef is, the smoother the resulting baseline

Type:: int

order_max

Maximum order of the baseline polynomial evaluation

Type:: int

distance

Minimum distance between baseline point to consider when doing automatic detection with ‘Spectrum.auto_baseline’

Type:: float

sigma

Smoothing coefficient (standard deviation) related to a gaussian filtering when defining baseline attached points to the spectrum

Type:: float

attached

Activation key for attach the baseline points to the spectrum

Type:: bool

is_subtracted

Key used to indicate whether the baseline has been subtracted from the spectrum

Type:: bool

y_eval

The baseline profile resulting from the ‘eval’ function

Type:: numpy.ndarray(n)

reinit(): Reinitialize the main attributes

add_point(x, y): Add point in the baseline

attached_points(x, y): Return baseline points attached to (x,y) ‘spectrum’ profile coords

load_baseline(fname): Load baseline from ‘fname’ with 1 header line and 2 (x,y) columns

eval(x, y, attached=False): Evaluate the baseline on a ‘x’ support and a ‘y’ profile possibly smoothed with a gaussian filter

plot(ax, x, y, attached=False, label='Baseline', show_all=True)

Plot the baseline and its related points

Parameters:

ax (Matplotlib.Axes) – Axis to work with
x (iterable of floats, optional) – Support to consider for the baseline plotting. If None, create a support from the baseline extrema points
y (iterable of floats, optional) – Values for baseline points attachment (if provided), sharing the same x coordinates
label (str, optional) – Label displays in the figure
show_all (bool, optional) – Activation key to display the primary baseline components (before attachment)

fitspy.baseline.arpls(y, coef=4, ratio=0.05, itermax=10)

Asymmetrically Reweighted Penalized Least Squares smoothing extracted from: https://irfpy.irf.se/projects/ica/_modules/irfpy/ica/baseline.html#arpls

Original article

Sung-June Baek, Aaron Park, Young-Jin Ahna and Jaebum Choo, Analyst, 2015, 140, 250 (2015), https://doi.org/10.1039/C4AN01061B

param y:: input data (i.e. spectrum intensity)
type y:: numpy.ndarray(n)
param coef:: parameter that can be adjusted by user. The larger coef is, the smoother the resulting background, y_smooth
type coef:: float, optional
param ratio:: wheighting deviations: 0 < ratio < 1, smaller values allow less negative values
type ratio:: float, optional
param itermax:: number of iterations to perform
type itermax:: int, optional
returns:: y_smooth – the fitted background
rtype:: numpy.ndarray(n)