pystoned 0.2.0

Estimating CNLS and StoNED in Python

StoNED-Python

StoNED-Python project provides the python codes for estimating Convex Nonparametric Least Square (CNLS), Stochastic Nonparametric Envelopment of Data (StoNED), and their different variants. It allows the user to estimate the CNLS/StoNED models in an open-access environment rather than in commercial software, e.g., GAMS, MATLAB. The StoNED-Python project is built based on the PYOMO.

Authors

• Timo Kuosmanen, Professor, Aalto University School of Business.
• Sheng Dai, Ph.D. candidate, Aalto University School of Business.

To do list

• CNLS/StoNED
  • variables returns to scale (VRS) model
  • constant returns to scale (CRS) model
  • Additive composite error term
  • Multiplicative composite error term
  • Residuals decomposition by method of moments(MoM)
  • Residuals decomposition by quasi-likelihood estimation(QLE)
  • Residuals decomposition by nonparametric kernel deconvolution (NKD)
• StoNEZD (contextual variables)
• Convex quantile regression (CQR)
• Convex expectile regression (CER)
• Isotonic CNLS (ICNLS)
• Isotonic convex quantile regression (ICQR)
• Isotonic convex expectile regression (ICER)
• Corrected convex nonparametric least squares (C²NLS)

Change log

[0.2.0] - 2020-04-19

Added

• CCNLS()
• CCNLS2()
• CNLSZ()

Changed

• Update REDAME.md
• Update function cqer()

[0.0.7] - 2020-04-18

Added

Changed

• Update function cnls()

Removed

[0.0.6] - 2020-04-17

Added

• README.md
• LICENSE.txt
• HISTORY.md

[0.0.2] - 2020-04-17

Added

• cqer()
• qllf()

[0.0.1] - 2020-04-01

Added

• stoned()
• cnls()