# load libraries
library(survey)
library(laeken)
# library( vardpoor )

# loop through epsilon values
for (this.epsilon in c(.5, 1, 2)) {
  # return test context
  context(paste(
    "gei epsilon =" ,
    this.epsilon ,
    "output survey.design and svyrep.design"
  ))

  ### test 1: test if funtion works on unweighted objects

  # load data
  data("api")

  # set up convey design
  expect_warning(dstrat1 <-
                   convey_prep(svydesign(id =  ~ 1, data = apistrat)))

  # perform tests
  test_that("svyatk works on unweighted designs", {
    expect_false(is.na (coef(
      svyatk(~ api00, design = dstrat1 , epsilon = this.epsilon)
    )))
    expect_false(is.na (SE(
      svyatk(~ api00, design = dstrat1 , epsilon = this.epsilon)
    )))
  })

  ### test 2: income data from eusilc --- data.frame-backed design object

  # collect and format data
  data(eusilc)
  names(eusilc) <- tolower(names(eusilc))

  # set up survey design objects
  des_eusilc <-
    svydesign(
      ids = ~ rb030 ,
      strata = ~ db040 ,
      weights = ~ rb050 ,
      data = eusilc
    )
  des_eusilc_rep <-
    as.svrepdesign(des_eusilc , type = "bootstrap" , replicates = 50)

  # prepare for convey
  des_eusilc <- convey_prep(des_eusilc)
  des_eusilc_rep <- convey_prep(des_eusilc_rep)

  # only striclty positive incomes
  test_that("error on income <= 0 " , expect_error(svyatk(~ eqincome , des_eusilc , epsilon = this.epsilon)))

  # filter positive
  des_eusilc <- subset(des_eusilc , eqincome > 0)
  des_eusilc_rep <- subset(des_eusilc_rep , eqincome > 0)

  # calculate estimates
  a1 <-
    svyatk(
      ~ eqincome ,
      des_eusilc ,
      epsilon = this.epsilon ,
      deff = TRUE ,
      linearized = TRUE ,
      influence = TRUE
    )
  a2 <-
    svyby(
      ~ eqincome ,
      ~ hsize,
      des_eusilc,
      svyatk ,
      epsilon = this.epsilon ,
      deff = TRUE ,
      covmat = TRUE ,
      influence = TRUE
    )
  a2.nocov <-
    svyby(
      ~ eqincome ,
      ~ hsize,
      des_eusilc,
      svyatk ,
      epsilon = this.epsilon ,
      deff = TRUE ,
      covmat = FALSE
    )
  b1 <-
    svyatk(
      ~ eqincome ,
      des_eusilc_rep ,
      epsilon = this.epsilon ,
      deff = TRUE ,
      linearized = TRUE
    )
  b2 <-
    svyby(
      ~ eqincome ,
      ~ hsize,
      des_eusilc_rep,
      svyatk ,
      epsilon = this.epsilon ,
      deff = TRUE ,
      covmat = TRUE
    )
  b2.nocov <-
    svyby(
      ~ eqincome ,
      ~ hsize,
      des_eusilc_rep,
      svyatk ,
      epsilon = this.epsilon ,
      deff = TRUE ,
      covmat = FALSE
    )

  # calculate auxilliary tests statistics
  cv_diff1 <- abs(cv(a1) - cv(b1))
  se_diff2 <- max(abs(SE(a2) - SE(b2)) , na.rm = TRUE)

  # perform tests
  test_that("output svyatk" , {
    expect_is(coef(a1) , "numeric")
    expect_is(coef(a2) , "numeric")
    expect_is(coef(b1) , "numeric")
    expect_is(coef(b2) , "numeric")
    expect_equal(coef(a1) , coef(b1))
    expect_equal(coef(a2) , coef(b2))
    expect_lte(cv_diff1 , coef(a1) * .20)         # the difference between CVs should be less than 5% of the coefficient, otherwise manually set it
    expect_lte(se_diff2 , max(coef(a2)) * .20)  # the difference between CVs should be less than 10% of the maximum coefficient, otherwise manually set it
    expect_is(SE(a1) , "matrix")
    expect_is(SE(a2) , "numeric")
    expect_is(SE(b1) , "numeric")
    expect_is(SE(b2) , "numeric")
    expect_lte(confint(a1)[1] ,  coef(a1))
    expect_gte(confint(a1)[2] , coef(a1))
    expect_lte(confint(b1)[, 1] , coef(b1))
    expect_gte(confint(b1)[2] , coef(b1))
    expect_equal(sum(confint(a2)[, 1] <= coef(a2)) , length(coef(a2)))
    expect_equal(sum(confint(a2)[, 2] >= coef(a2)) , length(coef(a2)))
    expect_equal(sum(confint(b2)[, 1] <= coef(b2)) , length(coef(b2)))
    expect_equal(sum(confint(b2)[, 2] >= coef(b2)) , length(coef(b2)))

    # check equality of linearized variables
    expect_equal(attr(a1 , "linearized") , attr(b1 , "linearized"))
    expect_equal(attr(a1 , "index") , attr(b1 , "index"))

    # check equality vcov diagonals
    expect_warning(expect_equal(diag(vcov(a2)) , diag(vcov(a2.nocov))))
    expect_warning(expect_equal(diag(vcov(b2)) , diag(vcov(b2.nocov))))

  })

  ### test 2: income data from eusilc --- database-backed design object

  # perform tests
  test_that("database svyatk", {
    # skip test on cran
    skip_on_cran()

    # load libraries
    library(RSQLite)
    library(DBI)

    # set-up database
    dbfile <- tempfile()
    conn <- dbConnect(RSQLite::SQLite() , dbfile)
    dbWriteTable(conn , 'eusilc' , eusilc)

    # database-backed design
    dbd_eusilc <-
      svydesign(
        ids = ~ rb030 ,
        strata = ~ db040 ,
        weights = ~ rb050 ,
        data = "eusilc",
        dbname = dbfile,
        dbtype = "SQLite"
      )

    # prepare for convey
    dbd_eusilc <- convey_prep(dbd_eusilc)

    # filter positive
    dbd_eusilc <- subset(dbd_eusilc , eqincome > 0)

    # calculate estimates
    c1 <-
      svyatk(
        ~ eqincome ,
        dbd_eusilc ,
        epsilon = this.epsilon ,
        deff = TRUE ,
        linearized = TRUE ,
        influence = TRUE
      )
    c2 <-
      svyby(
        ~ eqincome ,
        ~ hsize,
        dbd_eusilc,
        svyatk ,
        epsilon = this.epsilon ,
        deff = TRUE ,
        covmat = TRUE ,
        influence = TRUE
      )

    # remove table and close connection to database
    dbRemoveTable(conn , 'eusilc')
    dbDisconnect(conn)

    # peform tests
    expect_equal(coef(a1) , coef(c1))
    expect_equal(coef(a2) , coef(c2))
    expect_equal(SE(a1) , SE(c1))
    expect_equal(SE(a2) , SE(c2))
    expect_equal(deff(a1) , deff(c1))
    expect_equal(deff(a2) , deff(c2))
    expect_equal(vcov(a1) , vcov(c1))
    expect_equal(vcov(a2) , vcov(c2))

    # test equality of linearized variables
    expect_equal(colSums(attr(a1 , "linearized")) , colSums(attr(c1 , "linearized")))
    expect_equal(attr(a2 , "linearized") , attr(c2 , "linearized"))
    expect_equal(colSums(attr(a1 , "influence")) , colSums(attr(c1 , "influence")))
    expect_equal(colSums(attr(a2 , "influence")) , colSums(attr(c2 , "influence")))
    # expect_equal(attr(a1 , "index") , attr(c1 , "index"))
    # expect_equal(attr(a2 , "index") , attr(c2 , "index"))

  })

  ### test 3: compare subsetted objects to svyby objects

  # calculate estimates
  sub_des <-
    svyatk(
      ~ eqincome ,
      design = subset(des_eusilc , hsize == 1) ,
      epsilon = this.epsilon ,
      deff = TRUE ,
      linearized = TRUE ,
      influence = TRUE
    )
  sby_des <-
    svyby(
      ~ eqincome,
      by = ~ hsize,
      design = des_eusilc,
      FUN = svyatk ,
      epsilon = this.epsilon ,
      deff = TRUE ,
      covmat = TRUE ,
      influence = TRUE
    )
  sub_rep <-
    svyatk(
      ~ eqincome ,
      design = subset(des_eusilc_rep , hsize == 1) ,
      epsilon = this.epsilon ,
      deff = TRUE ,
      linearized = TRUE
    )
  sby_rep <-
    svyby(
      ~ eqincome,
      by = ~ hsize,
      design = des_eusilc_rep,
      FUN = svyatk ,
      epsilon = this.epsilon ,
      deff = TRUE ,
      covmat = TRUE
    )

  # perform tests
  test_that("subsets equal svyby", {
    # domain vs svyby: coefficients must be equal
    expect_equal(as.numeric(coef(sub_des)) , as.numeric(coef(sby_des[1, ])))
    expect_equal(as.numeric(coef(sub_rep)) , as.numeric(coef(sby_rep[1, ])))

    # domain vs svyby: SEs must be equal
    expect_equal(as.numeric(SE(sub_des)) , as.numeric(SE(sby_des[1, ])))
    expect_equal(as.numeric(SE(sub_rep)) , as.numeric(SE(sby_rep[1, ])))

    # domain vs svyby and svydesign vs svyrepdesign:
    # coefficients should match across svydesign
    expect_equal(as.numeric(coef(sub_des)) , as.numeric(coef(sby_rep[1, ])))

    # domain vs svyby and svydesign vs svyrepdesign:
    # coefficients of variation should be within five percent
    cv_diff <- max(abs(cv(sub_des) - cv(sby_rep)[1]))
    expect_lte(cv_diff , .05)

    # check equality of linearized variables
    expect_equal(attr(sub_des , "linearized") , attr(sub_rep , "linearized"))

    # check equality of variances
    expect_equal(vcov(sub_des)[1] , vcov(sby_des)[1, 1])
    expect_equal(vcov(sub_rep)[1] , vcov(sby_rep)[1, 1])

  })

  ### test 4: compare subsetted objects to svyby objects

  # compare database-backed designs to non-database-backed designs
  test_that("dbi subsets equal non-dbi subsets", {
    # skip test on cran
    skip_on_cran()

    # load libraries
    library(RSQLite)
    library(DBI)

    # set up database
    dbfile <- tempfile()
    conn <- dbConnect(RSQLite::SQLite() , dbfile)
    dbWriteTable(conn , 'eusilc' , eusilc)

    # create database-backed design (with survey design information)
    dbd_eusilc <-
      svydesign(
        ids = ~ rb030 ,
        strata = ~ db040 ,
        weights = ~ rb050 ,
        data = "eusilc",
        dbname = dbfile,
        dbtype = "SQLite"
      )

    # create a hacky database-backed svrepdesign object
    # mirroring des_eusilc_rep
    dbd_eusilc_rep <-
      svrepdesign(
        weights = ~ rb050,
        repweights = attr(des_eusilc_rep , "full_design")$repweights ,
        scale = attr(des_eusilc_rep , "full_design")$scale ,
        rscales = attr(des_eusilc_rep , "full_design")$rscales ,
        type = "bootstrap" ,
        data = "eusilc" ,
        dbtype = "SQLite" ,
        dbname = dbfile ,
        combined.weights = FALSE
      )

    # prepare for convey
    dbd_eusilc <- convey_prep(dbd_eusilc)
    dbd_eusilc_rep <- convey_prep(dbd_eusilc_rep)

    # filter positive incomes
    dbd_eusilc <- subset(dbd_eusilc , eqincome > 0)
    dbd_eusilc_rep <- subset(dbd_eusilc_rep , eqincome > 0)

    # calculate estimates
    sub_dbd <-
      svyatk(
        ~ eqincome ,
        design = subset(dbd_eusilc     , hsize == 1) ,
        epsilon = this.epsilon ,
        deff = TRUE ,
        linearized = TRUE ,
        influence = TRUE
      )
    sub_dbr <-
      svyatk(
        ~ eqincome ,
        design = subset(dbd_eusilc_rep , hsize == 1) ,
        epsilon = this.epsilon ,
        deff = TRUE ,
        linearized = TRUE
      )
    sby_dbd <-
      svyby(
        ~ eqincome,
        by = ~ hsize,
        design = dbd_eusilc     ,
        FUN = svyatk ,
        epsilon = this.epsilon ,
        deff = TRUE ,
        covmat = TRUE ,
        influence = TRUE
      )
    sby_dbr <-
      svyby(
        ~ eqincome,
        by = ~ hsize,
        design = dbd_eusilc_rep ,
        FUN = svyatk ,
        epsilon = this.epsilon ,
        deff = TRUE ,
        covmat = TRUE
      )

    # remove table and disconnect from database
    dbRemoveTable(conn , 'eusilc')
    dbDisconnect(conn)

    # perform tests
    expect_equal(coef(sub_des) , coef(sub_dbd))
    expect_equal(coef(sub_rep) , coef(sub_dbr))
    expect_equal(SE(sub_des) , SE(sub_dbd))
    expect_equal(SE(sub_rep) , SE(sub_dbr))
    expect_equal(deff(sub_des) , deff(sub_dbd))
    expect_equal(deff(sub_rep) , deff(sub_dbr))
    expect_equal(vcov(sub_des) , vcov(sub_dbd))
    expect_equal(vcov(sub_rep) , vcov(sub_dbr))

    # compare database-backed subsetted objects to database-backed svyby objects
    # dbi subsets equal dbi svyby
    expect_equal(as.numeric(coef(sub_dbd)) , as.numeric(coef(sby_dbd[1, ])))
    expect_equal(as.numeric(coef(sub_dbr)) , as.numeric(coef(sby_dbr[1, ])))
    expect_equal(as.numeric(SE(sub_dbd)) , as.numeric(SE(sby_dbd[1, ])))
    expect_equal(as.numeric(SE(sub_dbr)) , as.numeric(SE(sby_dbr[1, ])))
    expect_equal(vcov(sub_dbd) , vcov(sub_des))
    expect_equal(vcov(sub_dbr) , vcov(sub_rep))

    # compare equality of linearized variables
    expect_equal(colSums(attr(sub_dbd , "linearized")) , colSums(attr(sub_dbr , "linearized")))
    expect_equal(colSums(attr(sub_dbd , "linearized")) , colSums(attr(sub_des , "linearized")))
    expect_equal(attr(sub_dbr , "linearized") , attr(sub_rep , "linearized"))

    # compare equality of indices
    # expect_equal(attr(sub_dbd , "index") , attr(sub_dbr , "index"))
    # expect_equal(attr(sub_dbd , "index") , attr(sub_des , "index"))
    expect_equal(attr(sub_dbr , "index") , attr(sub_rep , "index"))

  })

}