Skip to main content

four component stochastic frontier model with determinants

Project description

SF4wD

four-component stochastic frontier model with determinants

Motivation

This package was developed to complement four-component stochastic frontier that consider

determinants in mean and variance parameters of inefficiency distributions

by Ruei-Chi Lee.

Installation

Install via $ pip install 4SFwD

Features

  • SF4wD: main.py - set method and model to run simulation or real data

  • HMC: Hamilton Monte Carlo designed for determinants parameters.

  • DA: Data augmentation for the model

  • TK: Two-parametrization method originally proposed by Tsiona and Kunmbhaker (2014) for four-component model without determinants.

  • PMCMC: Particle MCMC for the model (perferred approach) - speed up by GPU parallel computation

Example

Here is how you run a simulation estimation for a four-component stochastic frontier model via PMCMC:

  • Parameter setting guideline in the SF4wD.py

  • Simulation data only offers stochastic frontier model that consider determinants in both mean and variance parameter of inefficiencies.

import SF4wD
#model:str - different way to consider determinants
#method:str - different Bayesian method to estimate the model
#data_name : str - simulation data or data in data/.
#S : int - MCMC length
#H : int - number of particles in PMCMC
#gpu: boolean - use parallel computation to run PMCMC
#save: boolean - save MCMC data
my_model = SF4wD(model = 'D', method = 'PMCMC', data_name ='',S=10, H=100, gpu=False, save=False)
my_model.run()

output:

                  mean     sd  hpd_3%  hpd_97%  mcse_mean  mcse_sd  ess_mean  ess_sd  ess_bulk  ess_tail  r_hat
beta0            2.412  0.093   2.318    2.555      0.046    0.035       4.0     4.0       7.0      10.0    NaN
beta1            1.078  0.074   0.977    1.242      0.023    0.017      10.0    10.0      10.0      10.0    NaN
xi0              0.580  0.043   0.531    0.652      0.014    0.011       9.0     9.0       8.0      10.0    NaN
xi1              0.694  0.127   0.479    0.867      0.073    0.058       3.0     3.0       3.0      10.0    NaN
delta0           0.141  0.072   0.013    0.273      0.023    0.019      10.0     8.0      10.0      10.0    NaN
delta1           0.774  0.137   0.620    0.984      0.079    0.063       3.0     3.0       3.0      10.0    NaN
z0              -0.461  0.716  -1.844    0.609      0.376    0.291       4.0     4.0       4.0      10.0    NaN
z1               2.728  0.889   1.268    3.941      0.459    0.354       4.0     4.0       4.0      10.0    NaN
gamma0           0.662  0.092   0.500    0.773      0.052    0.041       3.0     3.0       3.0      10.0    NaN
gamma1           0.412  0.061   0.349    0.519      0.021    0.015       9.0     9.0       9.0      10.0    NaN
sigma_alpha_sqr  1.377  0.178   1.095    1.693      0.075    0.057       6.0     6.0       6.0      10.0    NaN
sigma_v_sqr      2.575  2.523   1.290    9.515      1.062    0.793       6.0     6.0       3.0      10.0    NaN

License

Ruei-Chi Lee is the main author and contributor.

Bug reports, feature requests, questions, rants, etc are welcome, preferably on the github page.

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

4SFwD-0.0.2.tar.gz (3.0 kB view details)

Uploaded Source

Built Distribution

4SFwD-0.0.2-py2.py3-none-any.whl (3.2 kB view details)

Uploaded Python 2 Python 3

File details

Details for the file 4SFwD-0.0.2.tar.gz.

File metadata

  • Download URL: 4SFwD-0.0.2.tar.gz
  • Upload date:
  • Size: 3.0 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/3.4.2 importlib_metadata/4.6.1 pkginfo/1.5.0.1 requests/2.22.0 requests-toolbelt/0.9.1 tqdm/4.40.2 CPython/3.7.3

File hashes

Hashes for 4SFwD-0.0.2.tar.gz
Algorithm Hash digest
SHA256 aab1f86812eb69105731bb8b119936d6e1c474ae1487f525d502318c49655772
MD5 7a4d78b89437de2229cee1b005835ded
BLAKE2b-256 fde7a95c4dee4bd283dd5c0e851bec437f3ff88f6eaaa8168c5c3fda5f5c1108

See more details on using hashes here.

File details

Details for the file 4SFwD-0.0.2-py2.py3-none-any.whl.

File metadata

  • Download URL: 4SFwD-0.0.2-py2.py3-none-any.whl
  • Upload date:
  • Size: 3.2 kB
  • Tags: Python 2, Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/3.4.2 importlib_metadata/4.6.1 pkginfo/1.5.0.1 requests/2.22.0 requests-toolbelt/0.9.1 tqdm/4.40.2 CPython/3.7.3

File hashes

Hashes for 4SFwD-0.0.2-py2.py3-none-any.whl
Algorithm Hash digest
SHA256 f7c65c24308fbc80b3861ef03743e1d1d0b5972c80152d85c241f1d00683a69b
MD5 cd6dc099e6d62fdce00c877eb5be63f7
BLAKE2b-256 78ae5726e55f42f483ea57eb5a70ce2201cd7646c4ff3ac8bf69a7b25ca39d19

See more details on using hashes here.

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page