Mateo Velásquez-Giraldo, Gustavo Canavire-Bacarreza, Kim P. Huynh, David 
T. Jacho-Chávez, "Flexible Estimation of Demand Systems: A Copula 
Approach", Journal of Applied Econometrics, Vol. 33, No. 7, 2018,
pp. 1009-1116.

The zip file vchj-code-data.zip contains three folders:

1. "data" contains tab-separated .txt files with demand data from
Chang & Serletis (2014).

2. "start_points" contains .Rdata files with estimates for every
model-household type combination under the assumption of trivariate
normality. These estimates are used as starting points in the
estimation of our alternate copula specifications.

3. "software" contains some R programs.


##########
## Data ##
##########

All data are taken from Chang & Serletis (2014)
http://qed.econ.queensu.ca/jae/2014-v29.2/chang-serletis/

o Files: Married_Couples_with_One_Child.txt
  Format: tab separated values

o Files: Married_Couples_without_Children.txt
  Format: tab separated values

o Files: Single_Members.txt
  Format: tab separated values

Note: No other data are used


#####################
# Note: replication #
#####################

Chang & Serletis (2014) assume trivariate normality of budget-share
residuals. We use estimates under the trivariate normality assumption
as starting points in the estimation of our alternative copula
specifications. Therefore, the script Gen_start_points.R can be used
to generate our estimates of the models considered by Chang & Serletis
(2014).


##############
## Software ##
##############

o Files: functions.R
  Format: .R file (source code with comments)
  Application: implements all the models and likelihood functions used
  in the paper.

o Files: Gen_start_points.R
  Format: .R file (source code with comments)
  Application: Generates/renews the starting points in the folder
  "start_points".
  Info: Uses all .csv data files above and functions.R.

o Files: Estimation_example.R
  Format: .R file (source code with comments)
  Application: estimates any of the demand system specifications mentioned
  in the article.
  Info: Uses all data files above and functions.R.


o The software packages used to obtain the results were:
  
- R language for statistical computing, version 3.4.4
  http://CRAN.R-project.org/

- bbmle 1.0.20
  https://cran.r-project.org/web/packages/bbmle/index.html

- matrixcalc 1.0-3
  https://cran.r-project.org/web/packages/matrixcalc/index.html

- numDeriv 2016.8-1
  https://cran.r-project.org/web/packages/numDeriv/index.html

- Matrix 1.2-12
  https://cran.r-project.org/web/packages/Matrix/index.html
     
- sn 1.5-1
  https://cran.r-project.org/web/packages/sn/index.html

- nloptr 1.0.4
  https://cran.r-project.org/web/packages/nloptr/index.html

- wrapr 1.4.0
  https://cran.r-project.org/web/packages/wrapr/index.html

- copula 0.999-18
  https://cran.r-project.org/web/packages/copula/index.html

- data.table 1.10.4-3
  https://cran.r-project.org/web/packages/data.table/index.html

- tictoc 1.0
  https://cran.r-project.org/web/packages/tictoc/index.html