The hardware and bandwidth for this mirror is donated by METANET, the Webhosting and Full Service-Cloud Provider.
If you wish to report a bug, or if you are interested in having us mirror your free-software or open-source project, please feel free to contact us at mirror[@]metanet.ch.
geex
A user had a case of estimating parameters based on a dataset that
contained only categorical predictors. The data can be represented
either as one row per individual or one row per group defined by unique
combinations of categories. In this example, I show how computations in
geex
can be massively sped up using the latter data
representation and the weights
option in
estimate_equation
.
The following code generates two datasets: data1
has one
row per unit and data2
has one row per unique combination
of the categorical varibles.
library(geex)
library(dplyr)
##
## Attaching package: 'dplyr'
## The following objects are masked from 'package:stats':
##
## filter, lag
## The following objects are masked from 'package:base':
##
## intersect, setdiff, setequal, union
set.seed(42)
<- 1000
n
<- data_frame(
data1 ID = 1:n,
Y_tau = rbinom(n,1,0.2),
S_star = rbinom(n,1,0.6),
Y = rbinom(n,1,0.4),
Z = rbinom(n,1,0.5))
## Warning: `data_frame()` was deprecated in tibble 1.1.0.
## Please use `tibble()` instead.
## This warning is displayed once every 8 hours.
## Call `lifecycle::last_lifecycle_warnings()` to see where this warning was generated.
<- data1 %>% group_by(Y_tau, S_star, Y, Z) %>% count() data2
This is the estimating equation that the user provided as an example. I have no idea what the target parameters represent, but it nicely illustrates the point.
<- function(data) {
example function(theta) {
with(data,
c( (1 - Y_tau)*(1 -Z )*(Y - theta[1]),
1-Y_tau)*Z*(Y-theta[2]),
(3] - theta[2]*theta[1]))
theta[
} }
The timing to find point and variance estimates is compared:
system.time({
<- m_estimate(
results1 estFUN = example,
data = data1,
root_control = setup_root_control(start = c(.5, .5, .5))
)})
## user system elapsed
## 0.526 0.002 0.528
system.time({
<- m_estimate(
results2 estFUN = example,
data = data2,
weights = data2$n,
root_control = setup_root_control(start = c(.5, .5, .5))
)})
## user system elapsed
## 0.036 0.003 0.040
The latter option is clearly preferred.
And the results are basically identical:
roots(results1)
## [1] 0.4123711 0.4014423 0.1655432
roots(results2)
## [1] 0.4123711 0.4014423 0.1655432
vcov(results1)
## [,1] [,2] [,3]
## [1,] 0.0006245391 0.0000000000 0.0002507164
## [2,] 0.0000000000 0.0005776115 0.0002381903
## [3,] 0.0002507164 0.0002381903 0.0001988710
vcov(results2)
## [,1] [,2] [,3]
## [1,] 6.245391e-04 6.873914e-47 0.0002507164
## [2,] 6.873914e-47 5.776115e-04 0.0002381903
## [3,] 2.507164e-04 2.381903e-04 0.0001988710
These binaries (installable software) and packages are in development.
They may not be fully stable and should be used with caution. We make no claims about them.