_images/logo.png

Elastic Functional Boxplots#

Elastic Functional Boxplots

moduleauthor:: J. Derek Tucker <jdtuck@sandia.gov>

class boxplots.ampbox(fdawarp)[source]#

This class provides amplitude boxplot for functional data using the SRVF framework

Usage: obj = ampbox(warp_data)

Parameters:
  • warp_data (fdawarp) – fdawarp class with alignment data

  • Q1 – First quartile

  • Q3 – Second quartile

  • Q1a – First quantile based on alpha

  • Q3a – Second quantile based on alpha

  • minn – minimum extreme function

  • maxx – maximum extreme function

  • outlier_index – indexes of outlier functions

  • f_median – median function

  • q_median – median srvf

  • plt – surface plot mesh

Author : J. D. Tucker (JDT) <jdtuck AT sandia.gov> Date : 15-Mar-2018

construct_boxplot(alpha=0.05, k_a=1)[source]#

This function constructs the amplitude boxplot using the elastic square-root slope (srsf) framework.

Parameters:
  • alpha – quantile value (e.g.,=.05, i.e., 95%)

  • k_a – scalar for outlier cutoff (e.g.,=1)

plot()[source]#

plot box plot and surface plot

Usage: obj.plot()

class boxplots.phbox(fdawarp)[source]#

This class provides phase boxplot for functional data using the SRVF framework

Usage: obj = phbox(warp_data)

Parameters:
  • warp_data (fdawarp) – fdawarp class with alignment data

  • Q1 – First quartile

  • Q3 – Second quartile

  • Q1a – First quantile based on alpha

  • Q3a – Second quantile based on alpha

  • minn – minimum extreme function

  • maxx – maximum extreme function

  • outlier_index – indexes of outlier functions

  • median_x – median warping function

  • psi_median – median srvf of warping function

  • plt – surface plot mesh

Author : J. D. Tucker (JDT) <jdtuck AT sandia.gov> Date : 15-Mar-2018

construct_boxplot(alpha=0.05, k_a=1)[source]#

This function constructs phase boxplot for functional data using the elastic square-root slope (srsf) framework.

Parameters:
  • alpha – quantile value (e.g.,=.05, i.e., 95%)

  • k_a – scalar for outlier cutoff (e.g.,=1)

plot()[source]#

plot box plot and surface plot

Usage: obj.plot()