-
Notifications
You must be signed in to change notification settings - Fork 0
/
DESCRIPTION
45 lines (45 loc) · 1.65 KB
/
DESCRIPTION
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
Package: lmabc
Title: Linear Regression with Abundance-Based Constraints (ABCs)
Version: 0.0.0.9000
Authors@R: c(
person("Dan", "Kowal", , "daniel.r.kowal@gmail.com", role = c("aut", "cre")),
person("Prayag", "Gordy", role = "aut"),
person("Virginia", "Baskin", role = "aut"),
person("Jai", "Uparkar", role = "aut"),
person("Caleb", "Fikes", role = "ctb")
)
Description: lmabc provides estimation and inference for linear regression
with categorical covariates (race, sex, etc.). Common strategies,
including the defaults in lm, select a "reference" group (e.g,. White,
Male, etc.) and present all results relative to this group. However,
this approach suffers from alarming biases, statistical
inefficiencies, and limited interpretability. lmabc uses
abundance-based constraints (ABCs) that (1) fully eliminate these
biases, (2) provide appealing statistical properties related to
estimation invariance and improved statistical power, and (3) produce
more interpretable model output. lmabc reproduces the core
functionality of lm and includes functions for generalized (logistic,
Poisson, etc.) and penalized (lasso, ridge, etc.) linear models.
Details are provided in Kowal (2024)
<doi:10.1101/2024.01.04.23300033>.
License: GPL (>= 3)
URL: https://drkowal.github.io/lmabc/,
https://github.com/drkowal/lmabc
Imports:
graphics,
stats,
utils
Suggests:
dplyr,
genlasso (>= 1.6.1),
glmnet (>= 4.0),
knitr,
rmarkdown,
testthat (>= 3.0.0),
tibble
VignetteBuilder:
knitr
Config/testthat/edition: 3
Encoding: UTF-8
Roxygen: list(markdown = TRUE)
RoxygenNote: 7.2.3