Skip to content

Latest commit

 

History

History
97 lines (81 loc) · 3.73 KB

classes_design.md

File metadata and controls

97 lines (81 loc) · 3.73 KB

Key words explainations

  • cases: different conditioned patterns / different visual intensity ratios
  • groups: fish groups trained with different experiment protocols (self-control group, experiment group, unpaired-control group)
  • classes: learners, non-learners, (all fish) which classified by the learning metrics (currently the significance test of the positional indices)

Loading/Saving functions

  • loadMats(byKeywords): append fishStack to an existing ABLITZER obj

    • ABLITZER method
    • arg1: keywords to filter files
    • arg2: append data / rewrite Data

  • loadYamls(byKeywords): import yaml files to matlab

    • ABLITZER method
    • arg1: keywords to filter files
    • arg2: append data / rewrite Data
    • arg3: old-flag to deal with old yaml files

  • saveData(by keywords/size): save fishStack in ABLITZER obj into multiples files

    • ABLITZER method
    • arg0: savingPath
    • arg1: which keywords to be classified
    • arg2: size to be divided

  • saveTrialsMat: save trial-by-trial mat in files

    • global function
    • structure needs to be defined

Utility functions (classification/finding/search for)

  • classifyFish： classify fish in groups by keys (e.g., ages, strains)

    • ABLITZER method

  • findFish: find specific fish by key-value pair or keywords

    • ABLITZER method

  • removeInvalidData: remove FISHDATA (in pair) whose data quality lower than the threshold

    • ABLITZER method

  • outputFeatures: output as print-out message/variable (struct)

    • ABLITZER method
    • arg1: indices to output (all/selective)

Analysis functions

  • evaluateDataQuality:
  • calcPItime(PositionalIndex):
    • FISHDATA method
    • phaseByPhase
      • add trials' values for each phase
      • add increments in RESDATA
  • calcPIturn(TurningIndex):
    • FISHDATA method
    • phaseByPhase
      • add trials' values for each phase
      • add increments in RESDATA
  • calcPIshock(ShockIndex):
    • FISHDATA method
    • phaseByPhase
      • add trials' values for each phase
  • [ifLearned, MemLen] = sayIfLearned
    • FISHDATA method
    • add "MemoryLength" in seconds
    • "ExtinctTime" should be converted to the idxFrame / seconds from the experiment beginning.

Plotting functions

  • plotPIs: plot performance index (positional/turning/shock) of different groups (1. exp only; 2. with self-control; 3. with unpaired control)

    • ABLITZER method
    • arg1: metric type (positional/turning/shock)
    • arg2: (number of arguments) which groups to plot (1. exp only; 2. with self-control; 3. with unpaired control)

  • plotDistance2centerline:

    • (FishData method)
    • arg1: the entire process / test-phase only / baseline-phase / training-phase
    • arg2: pixelSize
    • arg3: w/o extinction point (blue triangle)
    • arg4: w/o shocking events (red filled-circles)
    • arg5: w/o shadows to demarcate consecutive phases

  • plotLearningCurves: plot performance index (positional/turning/shock) of different classes (1. learners only; 2. with non-learners; 3. with all)

    • ABLITZER method
    • arg1: metric type (positional/turning/shock)
    • arg2: which classes to plot (1. learners only; 2. with non-learners; 3. with all)
    • we can also design a global function to deal with saved trialBytrialPImat.

  • histPlotMemLen: plot the histograms of memory lengths of different cases

    • ABLITZER method
    • arg1: which case(s) to plot. (white-black checkerboard, red-black checkerboard, pure-black patterns)
    • arg2: binEdges/width

  • plotOntogeny: plot different metrics (memory length, positional index increments, turning index increments) versus ages.

    • ABLITZER method
    • arg1: ages

Demo scripts

  • plot Figures 1-4
  • basic usages
    • arg2: corresponding indices to ages
    • arg3: metrics (memory length, positional index increments, turning index increments)