iq_readout.classifiers.TwoStateLinearClassifier

class TwoStateLinearClassifier(params)

Template for creating two-state linear classifiers.

The elements to be rewritten for each specific classifier are:

• _pdf_func_..., which specify the PDFs

• _pdf_func_..._proj, which specify the PDFs for the projected data

• _param_names, which specify the parameter names of the PDFs

• _param_names_proj, which specify the parameter names of the PDFs for the projected data.

• statistics, which computes the relevant statistics

• fit, which performs the fit

NB: if the classifier does not use max-likelihood classification, then predict needs to the overwritten.

__init__(params)

Loads params to this TwoStateLinearClassifier.

Parameters:

params

The structure of the dictionary must be

{
    0: {"param1": float, ...},
    1: {"param1": float, ...}
}

Methods

fit(shots_0, shots_1, **kargs)	Runs fit to the given data.
from_yaml(filename)	Load the TwoStateClassifier from YAML file.
pdf_0(z)	Returns \(p(z|0)\).
pdf_0_projected(z_proj)	Returns \(p_{proj}(z_{proj}|0)\).
pdf_1(z)	Returns \(p(z|1)\).
pdf_1_projected(z_proj)	Returns \(p_{proj}(z_{proj}|1)\).
predict(z[, p_0])	Classifies the given data to 0 or 1 using maximum-likelihood classification, which is defined by
project(z)	Returns the projection of the given IQ data to the \(\mu_0 - \mu_1\) axis.
to_yaml(filename)	Stores parameters in a YAML file.

Attributes

params	Returns the parameters required to set up the classifier.
params_proj	Returns the parameters for the projected PDFs, computed from params.
statistics	Returns dictionary with general statistical data: