AdaptiveETBayes

Code related to Adaptive Runtime Estimate of Task Execution Times using Bayesian Modeling

R-related version info

RStudio Version 1.3.1093

R version 4.0.3 (2020-10-10)

platform x86_64-w64-mingw32
arch x86_64
os mingw32
system x86_64, mingw32
svn rev 79318
nickname Bunny-Wunnies Freak Out
packages: depmixS4 version 1.4-2 data.tree version 1.0.0 MASS version 7.3.53 ggplot version 3.3.2

Python-related info

Python and GpyOpt used for preprocessing step 2 (find points of model change)

Install anaconda.

In anaconda prompt:

>conda update -y anaconda 
>conda update -y numpy scipy matplotlib
>conda update -y jupyter

Install environment python_3_gpyopt with dependencies:

conda env create -f environment.yml

Generate simulated sequences

Source from RStudio in AdaptiveETBayes project

createEvalSequences/simulateMarkovClusterSequences.R

Result found in data/simulatedSequences

Preprocessing step 1: Fit HMM

Run from RStudio in AdaptiveETBayes project

preprocessing1fitHMM/simMarkovFitHMMSequences.R

Result found in data/fittedHMMSequences

Preprocessing step 2: Find points of model change

In anaconda prompt:

conda activate python_3_gpyopt
cd preprocessing2pointsModelChange
python BOSplitTraceSimulatedBayesianSequences.py

Result found in data/splitIndicesSequencesPreprocess

Preprocessing step 3: Cluster segments + adaptive step

Adaptive step, full process (FP)

Source from RStudio in AdaptiveETBayes project

preprocessing3clusterAdaptive/clusterSequencesGLRSlidingWindowFP.R

Result found in data/resultsFP

Adaptive step, no create/ merge (NCM)

Source from RStudio in AdaptiveETBayes project

preprocessing3clusterAdaptive/clusterSequencesGLRSlidingWindowNCM.R

Result found in data/resultsNCM

Adaptive step, switch preprocessing (SP)

Source from RStudio in AdaptiveETBayes project

preprocessing3clusterAdaptive/clusterSequencesGLRSlidingWindowSP.R

Result found in data/resultsSP

Generate sequential lines for ground truth and estimated distributions (Fig 3)

Source from RStudio in AdaptiveETBayes project

evaluationFigures/exportDistrLines.R

Result found in

data/toLatexDistrLinesFP

data/toLatexDistrLinesNCM

data/toLatexDistrLinesSP

Calculate KL divergence and generate lines for graphs (Fig 3)

Source from RStudio in AdaptiveETBayes project

evaluationFigures/KLDivCalcSequence.R

Result found in

data/toLatexKLDiv