tests/testthat/test-test_logical_euclid_join.R

In zoomerjoin: Superlatively Fast Fuzzy Joins

test_that("euclidean_join_core works on toy datasets", {
  capture_messages({
    n <- 2000
    X_1 <- matrix(c(seq(0, 1, 1 / (n - 1)), seq(0, 1, 1 / (n - 1))), nrow = n)
    X_2 <- X_1 + .000000000001
    X_2 <- rbind(X_2, matrix(rep(c(2, 2), 10), nrow = 10))

    X_1 <- as.data.frame(X_1)
    X_2 <- as.data.frame(X_2)

    X_1$id_1 <- 1:n
    X_2$id_2 <- 1:(n + 10)

    X_1 <- as.data.frame(X_1)
    X_2 <- as.data.frame(X_2)

    inner_join_out <- euclidean_inner_join(X_1, X_2, threshold = .00005)

    expect_equal(nrow(inner_join_out), 2000)
    expect_true(all(inner_join_out$id_1 %in% 1:2000))
    expect_true(all(inner_join_out$id_2 %in% 1:2000))
    expect_true(all(inner_join_out$id_1 == inner_join_out$id_2))

    left_join_out <- euclidean_left_join(X_1, X_2, threshold = .000005, band_width = 1)


    expect_true(all(left_join_out$id_1 %in% 1:2000))
    expect_true(all(left_join_out$id_2 %in% 1:2000))
    expect_true(all(left_join_out$id_1 == left_join_out$id_2))
    expect_equal(nrow(left_join_out), 2000)

    right_join_out <- euclidean_right_join(X_1, X_2, threshold = .00005)
    expect_equal(nrow(right_join_out), 2010)
    expect_true(all(right_join_out$id_1 %in% c(1:2000, NA), na.rm = T))
    expect_true(all(right_join_out$id_2 %in% 1:2010))

    outer_join_out <- euclidean_anti_join(X_1, X_2, threshold = .00005)
    expect_equal(nrow(outer_join_out), 10)
    expect_true(all(outer_join_out$id_2 %in% 2000:2010))
    expect_true(all(is.na(outer_join_out$id_1)))

    full_join_out <- euclidean_full_join(X_1, X_2, threshold = .00005)
    expect_equal(nrow(full_join_out), 2010)
    expect_true(all(full_join_out$id_2 %in% 1:2010))
    expect_true(all(full_join_out$id_1 %in% c(1:2000, NA)))
  })
})


test_that("set.seed works with Euclidean join", {
  set.seed(1)
  n <- 20
  X_1 <- matrix(c(seq(0, 1, 1 / (n - 1)), seq(0, 1, 1 / (n - 1))), nrow = n)
  X_2 <- X_1 + .1
  X_2 <- rbind(X_2, matrix(rep(c(2, 2), 10), nrow = 10))
  X_1 <- as.data.frame(X_1)
  X_2 <- as.data.frame(X_2)
  X_1$id_1 <- 1:n
  X_2$id_2 <- 1:(n + 10)
  X_1 <- as.data.frame(X_1)
  X_2 <- as.data.frame(X_2)

  for (i in 1:20) {
    set.seed(i)
    suppressWarnings(out_1 <- euclidean_inner_join(X_1, X_2, threshold = 5))
    set.seed(i)
    suppressWarnings(out_2 <- euclidean_inner_join(X_1, X_2, threshold = 5))
    expect_equal(colSums(out_1), colSums(out_2))
  }
})