This repo contains all final versions of simulations and Rworkspace. Note that the source codes are only used for internal simulations and are not polished well to the latest version.
- homo_case.R: investigates settings S1 to S5 when homo errors.
- hetero_case.R: investigates settings S1 to S5 when hetero errors.
- homo_case_S2.R: investigates varying violation strengths (
$\tau$ ) for S2 when homo errors. - hetero_case_S2.R: investigates varying violation strengths (
$\tau$ ) for S2 when hetero errors.
- simu_illustrate_LUa.R: illustrates varying initial intervals and grid-size (LUa) effect on Searching and Sampling for S1 and S2 when homo errors.
- simu_illustrate_Lua_additional.R: Specifically for Searching method, it adds a few more options on grid-size based on the above one.
- simu_illustrate_S0.R: investigates the setting S0 for varying thresholds.
- simu_illustrate_samplingLambda.R: illustrates the effect of varying proportions on Sampling method for settings S1 to S5
- simu_illustrate_samplingM.R: illustrates the effect of sampling times M on Sampling method for settings S1 and S2
- simu_bootstrap.R: compares the bootstrap threshold vs. regular threshold for Searching method for settings S1 to S5
- simu_filtering.R: compares filtering vs. non-filtering effects on Sampling method for settings S1 to S5
- simu_highd_pz100.R: investigates SearchingSampling on high-d S2 setting when
$p_z=100$ - simu_highd_pz200.R: investigates SearchingSampling on high-d S2 setting when
$p_z=200$ - simu_CIIV.R: investigates SearchingSampling on settings CIIV-1 and CIIV-2
- simu_threshold_S2.R: computes the
$max_{T_{j,k}}$ for setting S2 - homo_S2_hists.R: get point estimators of TSHT and CIIV method for setting S2, prepared for later histograms
- summary.R: For each task, it reads the relevant RDatas and either outputs latex tables or figures as demanded.
- RDatas: the file contains all Rworkspace.
- RealData: contains Econ datasets and R files to clean and compare all methods on it.