Displaying 5 results from an estimated 5 matches for "mcreg".
Did you mean:
creg
2009 Dec 03
0
Problem with predict() and factors
...ing)")
#Ok our ranges are comfirmed we'll get a better graph later
# This Heavy Voodoo™ allows us to have a dynamic number of
#dummy variables we can add\remove from the spreadsheet
forecastDistance <- 52
#Grab Existing Regressors (clipping out the data)
cReg <- sh2[1:modLength,-1]
mcReg <- sh2[1:modMax,-1]
#transform the on\offs into proper factors
for(i in names(cReg)) cReg[[i]] <- factor(cReg[[i]])
for(i in names(mcReg)) mcReg[[i]] <- factor(mcReg[[i]])
#Grab X Future Regressors equal to the forecastDistance (gotta double check
if I need a +1 on the start point)
fReg &l...
2009 Jul 15
2
Spaces in a name
I am reading regressors from an excel file (I have no control over the file)
and some of the element names have spaces:
i.e. "Small Bank Aquired"
but I have found that lm(SourceData ~ . - "Small Bank Aquired", mcReg)
doesn't work (mcReg = modelCurrentRegressors)
As they are toggles I have ran them through factor() to be treated propertly
as 0 or 1 but due to the fact I am grabbing automagically the first 2/3rds
of the data some of the regressors are either all 0s or all 1s accordingly
so I need to take th...
2009 Jul 15
2
storing lm() results and other objects in a list
to clean up some code I would like to make a list of arbitrary length
to store?various objects for use in a loop
sample code:
############ BEGIN SAMPLE ##############
# You can see the need for a loop already
linearModel1=lm(modelSource ~ .,mcReg)
linearModel2=step(linearModel1)
linearModel3=lm(modelSource ~ .-1,mcReg)
linearModel4=step(linearModel3)
#custom
linearModel5=lm(modelSource ~ . -ACF-MonthlyST1-MonthlyST2-MonthlyBLA,mcReg)
LinearModel1.res <- residuals(linearModel1)
LinearModel2.res <- residuals(linearModel2)
LinearModel3....
2012 Jan 03
1
returning information from functions via attributes rather than return list
...ork to write the new information as object
attributes, and it seems to work. There is a generic function
"meanCenter" and a method "meanCenter.default". At the end of
meanCenter.default, here's my use (or abuse) of attributes.
res <- eval(mc)
class(res) <- c("mcreg", class(model))
attr(res, "centeredVars") <- nc
attr(res, "centerCall") <- match.call()
res
I wrote print and summary methods, but other methods that work for lm
objects like plot will also work for these new ones.
meanCenter <- function(model, centerOnl...
2009 Jul 21
2
Odd coefficent behavior
...TBUILD + UNITDB + ITBUILD + ITDB +
UATBUILD + UATDB + HOGANCODE + RCF + ReleaseST1 + ReleaseST2 +
ReleaseBLA + Small.Bank.Acquisitions + HLY.NewYear + HLY.MLK +
HLY.PRES + HLY.MEMORIAL + HLY.J4 + HLY.LABOR + HLY.COLUMBUS +
HLY.VETS + HLY.THANKS + HLY.XMAS + HLY.ELECT + HLY.PATRIOT + EOM,
data = mcReg)
Coefficients:
(Intercept) UNITBUILD1 UNITDB1
405.8326 -8.5675 13.5029
ITBUILD1 ITDB1 UATBUILD1
33.0950 -6.1938...