Skip to content

Latest commit

 

History

History
49 lines (40 loc) · 2.13 KB

developmentChangelog.md

File metadata and controls

49 lines (40 loc) · 2.13 KB

Development Changelog for AutoSMP

02.06.2021

  • Change Name of Mainclass from TWiseParameterSampler to AutoSMP
  • Change package structure of JS implementation to de.tubs.cs.isf.AutoSMP.
  • Change the input parameter of Mainclass to take configPath and configName as first parameters
  • Refactor comments for properties

Todo 03.06.2021

[x] implement pathproperty and set defaults [x] extend config file with implemented path properties [x] DEFAULT_LOG_DIRECTORY to be created [x] DEFAULT_CSV_DIRECTORY to be created [x] create DEFAULT_TEMP_DIRECTORY [] fix sampling stability calculation

03.06.2021

  • extend original config.properties from JS
  • change path setting in SamplingConfig/readConfigFile to directory properties set by properties file
  • remove stability calculation from AutoSMP temporary
  • change refreshPaths method in SamplingConfig to use property values

09.06.2021

  • there must be a sample.csv file present in your temp directory otherwise fide throws an error

Todos

  • Auftrennen der Model Informationen und der sampling-evaluation results
  • Remove Author field from evaluation results

16.06.2021

  • Chvatal algorithm implemented and running
  • reading sample size, sample time and coverage works
  • storing sample does not work, needs re-implementation
  • memory analysis needs to be removed from csv files
  • evalautaion result csv append does not work, needs to be checked
  • validity check does not work --> confusion of cnf formulas

18.06.2021

  • remove memory information from out out csv (WriterModule line 138)
  • reduce sampling stability to one metric in csv header (writer_module line 46 to 54)

21.06.2021

  • fix bug Null Pointer when removing unselected features from sample during stability calculation
  • fix bug more than one default value for missing stability is writen
  • adjust sampling interfaces to take maxSize and randomSeed as parameters
  • make algorithm tool path more dynamic by introducing a parameter in the adapter interfaces
  • tidy-up adapter interfaces by introducing a new project specially for this purpose
  • implement sampling algorithms (Chvatal, ICPL, YASA, IncLing, RandomSamp) in separate project