Comments (5)
This seems a little tricky, as we want to pass trace=FALSE
to some models but not others. I see 2 options:
- Provide an optional list of arguments for the models, e.g.
list(nn=list(trace=FALSE), gbm=list(verbose=FALSE))
- Make methodList an actual list, e.g.
list(glm=NULL, rpart=NULL, nn=list(trace=FALSE), gbm=list(verbose=FALSE))
If we go with #1, we can do it post-CRAN release. If we go with #2, I think it should be pre-release.
from caretensemble.
Here's another option that might work, post CRAN release: a methodList class. If a character vector is passed to buildModels, it coerces it to a methodList. The methodList class would be constructed from objects like this:
list(
"glm",
"rpart",
"nnet"=list(trace=FALSE),
"gbm"=list(verbose=FALSE)
)
And would return an object like this:
list(
"glm"=NULL,
"rpart"=NULL,
"nnet"=list(trace=FALSE),
"gbm"=list(verbose=FALSE)
)
In other words, un-named objects in the input list would be ok so long as they were character vectors, and they would be converted to the "model_function"=model_arguments format. The modelList generator could also use modelLookup() from caret to verify that all models are valid.
I think we can safely push this post 1.0, maybe targeting 1.1 or something.
from caretensemble.
Yes. This sounds like a great solution.
from caretensemble.
Ok. We'll put that on the agenda for post-1.0.0
from caretensemble.
Closed by #56
from caretensemble.
Related Issues (20)
- Fix travis build
- print.caretStack is hardcoded to say "ensemble of 2 base models" HOT 1
- could not find function "residuals.caretEnsemble" HOT 6
- ClassPredictions not saving HOT 7
- Adaptive LASSO with caretEnsemble HOT 5
- Fix CRAN build HOT 1
- CaretEnsemble with pretrained models HOT 2
- Allow predict without newdata HOT 1
- Add commands workflow HOT 1
- Cancel previous workflow on new commit HOT 1
- run lint job first, then unit tests, then R CMD CHECK HOT 1
- 100% code coverage
- caretList improvement HOT 1
- remove caretEnsemble completely HOT 1
- Review all skip_on_cran() testsβ some of them maybe shouldn't be skipped
- Replace all suppressWarnings with expect_warnings in tests
- Big Data testing
- Remove tuneList and only use methodList HOT 1
- add defaultTrainControl, use it in caretList and caretStack HOT 1
- Prep for CRAN release HOT 1
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
π Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. πππ
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google β€οΈ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from caretensemble.