Sokbae Lee, Ryo Okui, and Yoon-Jae Whang, "Doubly Robust Uniform 
Confidence Band for the Conditional Average Treatment Effect Function", 
Journal of Applied Econometrics, Vol. 32, No. 7, 2017, pp. 1207-1225.

All R program files are zipped in low-programs.zip. These are ASCII
files in DOS format, so Unix users should use "unzip -a". The data
files are zipped in low-data.zip. The text files are ASCII files in
DOS format, but one file is a binary Stata .dta file.

1. Monte Carlo results

   a. "drcate_monte.R": main file
   b. "gdp.R": function to generate data
   c. "getpsi.R": function to compute the "Psi" function in the text
   d. "critical_value.R": function to compute the critical values

   These codes require the following two packages: "locpol" and
   "KernSmooth". The results presented in the paper are computed with
   R 3.3.1 with Windows 10. 

   To obtain the Monte Carlo results, put the four files listed above
   to a folder, change
   "/Users/okuiryo/project/drcate/replication_files" in line 9 of
   "drcate_monte.R" to the name of folder that contains these four
   files, and run "drcate_monte.R". The results are saved in
   "results.txt", "result_est.csv" and "result_ci.csv"

2. Empirical Application 1: Pennsylvania data

   a. "drcate_application_pn.R": main file
   
   These codes require the following three packages: "readstata13",
   "locpol" and "KernSmooth". The results presented in the paper are
   computed with R 3.3.1 with Windows 10. 
   
   The data file is available at
   "http://www.stata-press.com/data/r13/cattaneo2.dta”

   To obtain the results in Figure 1, put the code and the data file
   to a folder, change
   "/Users/okuiryo/project/drcate/replication_files" in line 10 of
   "drcate_application_pn.R" to the name of folder that contains these
   two files, and run "drcate_application_pn.R". The results are saved
   in "bw_cb5p.pdf". Note that running the codes also produces
   "bw_cp1p.pdf" and "bw_cp10p.pdf" in which the confidence levels are
   set to 1% and 10%, respectively.


3. Empirical Application 2: North Carolina data

   a. "drcate_application_cn.R": main file for Figure 2
   b. "drcate_application_cn_oldcov.R": main file for Figure 3
   
   These codes require the following three packages: "foreign",
   "locpol" and "KernSmooth". The results presented in the paper are
   computed with R 3.3.1 with Windows 10. 
   
   The data file is available at
   “http://www.personal.ceu.hu/staff/Robert_Lieli/cate-birthdata.zip"
   We use the file "cate-birthdata-cleaned-1stkid-white-detailed.txt".

   To obtain the results in Figure 2, put "drcate_application_pn.R"
   and the data file to a folder, change
   "/Users/okuiryo/project/drcate/replication_files" in line 10 of
   "drcate_application_cn.R" to the name of folder that contains these
   two files, and run "drcate_application_cn.R". The results are saved
   in "bw_cb5p_white.pdf". Note that running the codes also produces
   "bw_cp1p_white.pdf" and "bw_cp10p_white.pdf" in which the
   confidence levels are set to 1% and 10%, respectively.

   To obtain the results in Figure 3, put
   "drcate_application_oldcov.R" and the data file to a folder, change
   "/Users/okuiryo/project/drcate/replication_files" in line 10 of
   "drcate_application_cn.R" to the name of folder that contain these
   two files, and run "drcate_application_cn_oldcov.R". The results
   are saved in "bw_cb5p_oldcov.pdf". Note that running the codes also
   produces "bw_cp1p_oldcov.pdf" and "bw_cp10p_oldcov.pdf" in which
   the confidence levels are set to 1% and 10%, respectively.