Error in riv package

On Aug 16, 2016, at 11:07 AM, Shivi Bhatia <shivipmp82 at gmail.com> wrote:

HI Team,

I am working to reduce the # of predictor variables from the model using
woe and iv values. For this i am using riv package. However i am having a
hard time installing this package:

install_github("riv","tomasgreif")
install.packages("DBI",dependencies=TRUE)

The error i receive is :
*Username parameter is deprecated. Please use tomasgreif/riv *

No. You did not get an error message. It was clearly labeled a :Warning
message". Here is the full console output from that call:

install_github("riv","tomasgreif")

Downloading GitHub repo tomasgreif/riv at master
from URL https://api.github.com/repos/tomasgreif/riv/zipball/master
Installing woe
'/Library/Frameworks/R.framework/Resources/bin/R' --no-site-file
--no-environ  \
  --no-save --no-restore --quiet CMD INSTALL  \
  '/private/var/folders/yq/m3j1jqtj6hq6s5mq_v0jn3s80000gn/T/Rtmp6YDYoj/
devtoolsadc32ead1a82/tomasgreif-woe-43fcf26'  \
  --library='/Library/Frameworks/R.framework/Versions/3.3/Resources/library'
\
  --install-tests

* installing *source* package ?woe? ...
** R
** data
*** moving datasets to lazyload DB
** demo
** preparing package for lazy loading
** help
*** installing help indices
** building package indices
** testing if installed package can be loaded
* DONE (woe)
Warning message:
Username parameter is deprecated. Please use tomasgreif/riv


It's just telling you to use this next time:

install_github("tomasgreif/riv")

I will lay long odds that you already have the package installed.

I cannot comment on what this package purports to do. I'm somewhat
suspicious that it is statistically suspect.

--
David.

The other approach i have to use the library(InformationValue). I have

this

as below:
WOE(X=SFDC1$support_cat, Y=SFDC1$survey)
WOETable(X=SFDC1$support_cat, Y=SFDC1$survey)
IV(X=SFDC1$support_cat, Y=SFDC1$survey)

This package assists me achieve what i am looking for but here i need to
add one independent variable at a time to see whether it is predictive or
not. Is there a way i can add all variables at a go or have to add one by
one,

For the above package (riv) i have seen an example which helps to take

the

entire data range and  predictive the power of each predictor. The link

for

the same is:
https://www.r-bloggers.com/r-credit-scoring-woe-

information-value-in-woe-

package/

Thanks, Shivi

      [[alternative HTML version deleted]]

posting-guide.html

David Winsemius
Alameda, CA, USA