linCombo: Estimate and CI for Linear Combination of Parameters

In s-huebler/regsisF20: MATH 5573 Functions

Description

Estimate and CI for Linear Combination of Parameters

Usage

linCombo(Y, X, a, l, alpha = 0.95)

Arguments

Y,	a vector of outputs
X	, a design matrix
a	, parameter values
l	, a number (estimate of output given parameters)
alpha	, confidence level

Value

, a confidence interval and pairs figure

Examples

clerical<-data.frame(Hours=c(128.5, 113.6, 146.6, 124.3, 100.4, 119.2, 109.5, 128.5, 131.2, 112.2),
Mail=c(7781,7004,7267,2129,4878,3999,11777,5764,7392,8100),
Gifts=c(100,110,61,102,45,144,123,78,172,126),
Charge=c(886,962,1342,1153,803,1127,627,748,876,685),
Returns=c(235,388,398,457,577,345,326,161,219,287),
Checks=c(644,589,1081,891,537,563,402,495,823,555)
)

lm1<-stats::lm(Hours~Mail+Gifts+Charge+Returns+Checks, data=clerical)
design<-stats::model.matrix(lm1)
a1<-matrix(c(0,1,-1,0,0,0), nrow=6)

coeff<-lm1$coefficients
l1<-coeff[2]-coeff[3]

linCombo(clerical$Hours, design, a1, l1 )

s-huebler/regsisF20 documentation built on Oct. 15, 2020, 10:42 p.m.