December 8, 2022
View c6ea2926

- randomSelection function added. - crissCrossValidate and crissCrossPlot contributions by Harry Robertson added and harmonised to ClassifyR code style. - selectionMethod and classifier defaults become "auto" for crossValidate. Previously only documented but not implemented in code. - predict method for standalone use now finds the correct prediction method for each trained model.

Dario Strbenac authored on 08/12/2022 06:15:07
November 29, 2022
View 5b348ed4

- Closing bracket in if check corrected.

Dario Strbenac authored on 29/11/2022 02:57:19
November 27, 2022
View acc7fad1

- Random forest's survival mode now calculates risk values by multiplying sum of survival probabilities by negative one. - performancePlot and samplesMetricMap now use a variable named metric to consistently refer to the metric. - samplesMetricMap now has default value of auto for metric. Also, the metric will be calculated rather then stopping with an error, if the metric hasn't been calculated previously using calcCVperformance. - samplesMetricMap default metric is accuracy, to be consistent with performancePlot.

Dario Strbenac authored on 27/11/2022 10:50:16
November 25, 2022
View b7f37b96

- performancePlot no longer ignores orderingList. - Default random forest function now seamlessly works when outcome is a Surv object so classifier doesn't need to be changed from default setting if input data is survival.

Dario Strbenac authored on 25/11/2022 11:40:05
View eb01a641

make makevars protable

Nick authored on 25/11/2022 04:12:52
View b95f2200

remove Sint from C++ code

Nick authored on 25/11/2022 04:12:43
View 1ffd6658

change to 2 workers

Nick authored on 25/11/2022 03:47:05
View a1de49b8

change to 2 workers

Nick authored on 25/11/2022 03:46:47
View 56367922

bump roxygen version

Nick authored on 25/11/2022 03:46:27
November 24, 2022
View b0b2eb70

ignore compiled binaries

Nick authored on 24/11/2022 00:44:32
View f62c2b4e

add makevar indicating c++11 compilation

Nick authored on 24/11/2022 00:28:27
November 22, 2022
View c2bb11be

- .doSelection modified so that feature selection doesn't happen twice for selectMulti function. - selectMulti fixed so correct indices are returned relative to the whole training data table, not to an individual assay table.

Dario Strbenac authored on 22/11/2022 04:52:30
November 17, 2022
View 1cd8cc5d

Merge pull request #53 from SydneyBioX/master

Dario Strbenac authored on 17/11/2022 09:43:39 • GitHub committed on 17/11/2022 09:43:39
View 2ad6f285

- Removed external documentation for internal utility functions.

Dario Strbenac authored on 17/11/2022 09:40:49
View e2dcdb98

- Corrected ranger random forest to do two model fits in each iteration; one for prediction and one for variable importance. Previously, a mode of model fitting which uses permuted features was being executed.

Dario Strbenac authored on 17/11/2022 05:55:04