openest.models.features_interpreter module

Probability Features File

The probability features file has the following format:

dpc1,<p-header-1>,<p-header-2>,...
<x-value-1>,g_1(y | x_1),g_2(y | x_1),...
<x-value-2>,g_1(y | x_2),g_2(y | x_2),...
...

<p-header> headers can be any of the following, with the corresponding values in their rows (<p-value-ij>).

• mean: E y|x_i
• var: E (y|x_i - E y|x_i)^2
• sdev: sqrt{E (y|x_i - E y|x_i)^2}
• skew: E ((y|x_i - E y|x_i) / sqrt{E (y|x_i - E y|x_i)^2})^3
• mode: max f(y | x_i)
• numeric (0 - 1): F^{-1}(p_j|x_i)

The row headers (<x-value>) can be numeric, in which case a continuous spline bridges them, or categorical strings.

Below is a sample features file:

dpc1,mean,var
treated,0,1
control,4,4

class openest.models.features_interpreter.FeaturesInterpreter

static best_knot(newknots)

Find the knot furthest from existing knots

static best_spline(row, limits)

static evaluate_spline(row, spline, limits)

static features_to_exponential(row, limits)

static features_to_gaussian(row, limits)

static features_to_uniform(row, limits)

static init_from_feature_file(file, delimiter, limits, status_callback=None)

static make_conditional(row, limits)

static make_conditional_respecting(row, limits)

static skew_gaussian_construct(lps, low_segment, high_segment)

static skew_gaussian_evaluate(lps, low_segment, high_segment, mean, lowp, highp)