get_lr: Do logistic regression to get a single ensemble member's 'a'...

In kdayday/forecasting: Probabilistic Solar Forecsting

Description

Do logistic regression to get a single ensemble member's 'a' coefficients. Uses a penalized/Firth linear regression to handle quasi- or complete separation Telemetry within given tolerance of 1 are assumed to be clipped glm and lm omit NA values by default

Usage

get_lr(fc, tel, form, A_transform, percent_clipping_threshold)

Arguments

fc	Vector of training forecast data on [0,1]
tel	Vector of telemetry data on [0,1]
form	Formula in terms of x,y for logistic regression model
A_transform	A function for pre-transforming the forecast data (optional), e.g. function(X) return(log(x)+1)
percent_clipping_threshold	Percentage threshold for determining if clipping is occurring

Value

Summary of the glm model

kdayday/forecasting documentation built on Oct. 7, 2020, 7:16 p.m.