Gilead-BioStats · Zhongkai-Wang · Jan 13, 2023 · Dec 9, 2022 · Jan 6, 2023 · Jan 6, 2023
diff --git a/inst/qualification/spec_QueryRate_Assess.yaml b/inst/qualification/spec_QueryRate_Assess.yaml
@@ -0,0 +1,36 @@
+- ID: SX_1
+Description: Given correct input data a Query Rate assessment can be done using the poisson method
+Risk: High
+Impact: High
+Tests:
+  - TX_X
+- ID: SX_2
+Description: Given correct input data a Query Rate assessment can be done using the identity method
+Risk: High
+Impact: High
+Tests:
+  - TX_X
+- ID: SX_3
+Description: Given correct input data a Query Rate assessment can be done using the normal approximation method
+Risk: High
+Impact: High
+Tests:
+  - TX_X
+- ID: SX_4
+Description: Assessments are correctly grouped by a grouping variable such as site or country for KRIs and study for QTLs when applicable
+Risk: Low
+Impact: High
+Tests:
+  - TX_X
+- ID: SX_5
+Description: Given correct input data flags will correctly be applied to records that meet flagging criteria, including custom thresholding
+Risk: Medium
+Impact: High
+Tests:
+  - TX_X
+- ID: S1_6
+Description: Given subset input data a Query Rate assessment can be done
+Risk: Medium
+Impact: Medium
+Tests:
+  - TX_X
diff --git a/tests/testqualification/qualification/test_qual_T10_1.R b/tests/testqualification/qualification/test_qual_T10_1.R
@@ -0,0 +1,45 @@
+test_that("Query rate assessment can return a correctly assess
10000
ed data frame for the identity test grouped by the site variable when given correct input data from clindata and the results should be flagged correctly using a custom threshold.", {
+  # gsm analysis
+  dfInput <- gsm::QueryRate_Map_Raw()
+
+  test10_1 <- QueryRate_Assess(
+    dfInput = dfInput,
+    strMethod = "Identity",
+    strGroup = "Site",
+    vThreshold = c(0.005, 0.05)
+  )
+
+  # double programming
+  t10_1_input <- dfInput
+
+  t10_1_transformed <- dfInput %>%
+    qualification_transform_counts(countCol = "Count",
+                                   exposureCol = "DataPoint")
+
+  t10_1_analyzed <- t10_1_transformed %>%
+    mutate(
+      Score = Metric
+    ) %>%
+    arrange(Score)
+
+  class(t10_1_analyzed) <- c("tbl_df", "tbl", "data.frame")
+
+  t10_1_flagged <- t10_1_analyzed %>%
+    qualification_flag_identity(threshold = c(0.005, 0.05))
+
+  t10_1_summary <- t10_1_flagged %>%
+    select(GroupID, Numerator, Denominator, Metric, Score, Flag) %>%
+    arrange(desc(abs(Metric))) %>%
+    arrange(match(Flag, c(2, -2, 1, -1, 0)))
+
+
+  t10_1 <- list(
+    "dfTransformed" = t10_1_transformed,
+    "dfAnalyzed" = t10_1_analyzed,
+    "dfFlagged" = t10_1_flagged,
+    "dfSummary" = t10_1_summary
+  )
+
+  # compare results
+  expect_equal(test10_1$lData, t10_1)
+})
diff --git a/tests/testqualification/qualification/test_qual_T10_2.R b/tests/testqualification/qualification/test_qual_T10_2.R
@@ -0,0 +1,49 @@
+test_that("Query rate assessment can return a correctly assessed data frame for the identity test grouped by a custom variable when given subset input data from clindata and the results should be flagged correctly.", {
+  # gsm analysis
+  dfInput <- gsm::QueryRate_Map_Raw(dfs = list(
+    dfQUERY = clindata::edc_queries %>% filter(foldername == "Week 120"),
+    dfSUBJ = clindata::rawplus_dm,
+    dfDATACHG = clindata::edc_data_change_rate
+  ))
+
+  test10_2 <- QueryRate_Assess(
+    dfInput = dfInput,
+    strMethod = "Identity",
+    strGroup = "CustomGroup"
+  )
+
+  # double programming
+  t10_2_input <- dfInput
+
+  t10_2_transformed <- dfInput %>%
+    qualification_transform_counts(countCol = "Count",
+                                   exposureCol = "DataPoint",
+                                   GroupID = "CustomGroupID")
+
+  t10_2_analyzed <- t10_2_transformed %>%
+    mutate(
+      Score = Metric
+    ) %>%
+    arrange(Score)
+
+  class(t10_2_analyzed) <- c("tbl_df", "tbl", "data.frame")
+
+  t10_2_flagged <- t10_2_analyzed %>%
+    qualification_flag_identity(threshold = c(0.00006, 0.01))
+
+  t10_2_summary <- t10_2_flagged %>%
+    select(GroupID, Numerator, Denominator, Metric, Score, Flag) %>%
+    arrange(desc(abs(Metric))) %>%
+    arrange(match(Flag, c(2, -2, 1, -1, 0)))
+
+
+  t10_2 <- list(
+    "dfTransformed" = t10_2_transformed,
+    "dfAnalyzed" = t10_2_analyzed,
+    "dfFlagged" = t10_2_flagged,
+    "dfSummary" = t10_2_summary
+  )
+
+  # compare results
+  expect_equal(test10_2$lData, t10_2)
+})
diff --git a/tests/testqualification/qualification/test_qual_T10_3.R b/tests/testqualification/qualification/test_qual_T10_3.R
@@ -0,0 +1,46 @@
+test_that("Query rate assessment can return a correctly assessed data frame for the identity test grouped by the country variable when given correct input data from clindata and the results should be flagged correctly.", {
+  # gsm analysis
+  dfInput <- gsm::QueryRate_Map_Raw()
+
+  test10_3 <- QueryRate_Assess(
+    dfInput = dfInput,
+    strMethod = "Identity",
+    strGroup = "Country"
+  )
+
+  # double programming
+  t10_3_input <- dfInput
+
+  t10_3_transformed <- dfInput %>%
+    qualification_transform_counts(countCol = "Count",
+                                   exposureCol = "DataPoint",
+                                   GroupID = "CountryID")
+
+  t10_3_analyzed <- t10_3_transformed %>%
+    mutate(
+      Score = Metric
+    ) %>%
+    arrange(Score)
+
+  class(t10_3_analyzed) <- c("tbl_df", "tbl", "data.frame")
+
+
+  t10_3_flagged <- t10_3_analyzed %>%
+    qualification_flag_identity(threshold = c(0.00006, 0.01))
+
+  t10_3_summary <- t10_3_flagged %>%
+    select(GroupID, Numerator, Denominator, Metric, Score, Flag) %>%
+    arrange(desc(abs(Metric))) %>%
+    arrange(match(Flag, c(2, -2, 1, -1, 0)))
+
+
+  t10_3 <- list(
+    "dfTransformed" = t10_3_transformed,
+    "dfAnalyzed" = t10_3_analyzed,
+    "dfFlagged" = t10_3_flagged,
+    "dfSummary" = t10_3_summary
+  )
+
+  # compare results
+  expect_equal(test10_3$lData, t10_3)
+})
diff --git a/tests/testqualification/qualification/test_qual_T10_4.R b/tests/testqualification/qualification/test_qual_T10_4.R
@@ -0,0 +1,42 @@
+test_that("Query rate assessment can return a correctly assessed data frame for the poisson test grouped by the site variable when given correct input data from clindata and the results should be flagged correctly using a custom threshold.", {
+  # gsm analysis
+  dfInput <- gsm::QueryRate_Map_Raw()
+
+  test10_4 <- QueryRate_Assess(
+    dfInput = dfInput,
+    strMethod = "Poisson",
+    strGroup = "Site",
+    vThreshold = c(-6, -4, 4, 6)
+  )
+
+  # double programming
+  t10_4_input <- dfInput
+
+  t10_4_transformed <- dfInput %>%
+    qualification_transform_counts(countCol = "Count",
+                                   exposureCol = "DataPoint")
+
+  t10_4_analyzed <- t10_4_transformed %>%
+    qualification_analyze_poisson()
+
+  class(t10_4_analyzed) <- c("tbl_df", "tbl", "data.frame")
+
+  t10_4_flagged <- t10_4_analyzed %>%
+    qualification_flag_poisson(threshold = c(-6, -4, 4, 6))
+
+  t10_4_summary <- t10_4_flagged %>%
+    select(GroupID, Numerator, Denominator, Metric, Score, Flag) %>%
+    arrange(desc(abs(Metric))) %>%
+    arrange(match(Flag, c(2, -2, 1, -1, 0)))
+
+
+  t10_4 <- list(
+    "dfTransformed" = t10_4_transformed,
+    "dfAnalyzed" = t10_4_analyzed,
+    "dfFlagged" = t10_4_flagged,
+    "dfSummary" = t10_4_summary
+  )
+
+  # compare results
+  expect_equal(test10_4$lData[names(test10_4$lData) != "dfBounds"], t10_4)
+})
diff --git a/tests/testqualification/qualification/test_qual_T10_5.R b/tests/testqualification/qualification/test_qual_T10_5.R
@@ -0,0 +1,46 @@
+test_that("Query rate assessment can return a correctly assessed data frame for the poisson test grouped by a custom variable when given subset input data from clindata and the results should be flagged correctly.", {
+  # gsm analysis
+  dfInput <- gsm::QueryRate_Map_Raw(dfs = list(
+    dfQUERY = clindata::edc_queries %>% filter(foldername == "Week 120"),
+    dfSUBJ = clindata::rawplus_dm,
+    dfDATACHG = clindata::edc_data_change_rate
+  ))
+
+  test10_5 <- QueryRate_Assess(
+    dfInput = dfInput,
+    strMethod = "Poisson",
+    strGroup = "CustomGroup"
+  )
+
+  # double programming
+  t10_5_input <- dfInput
+
+  t10_5_transformed <- dfInput %>%
+    qualification_transform_counts(countCol = "Count",
+                                   exposureCol = "DataPoint",
+                                   GroupID = "CustomGroupID")
+
+  t10_5_analyzed <- t10_5_transformed %>%
+    qualification_analyze_poisson()
+
+  class(t10_5_analyzed) <- c("tbl_df", "tbl", "data.frame")
+
+  t10_5_flagged <- t10_5_analyzed %>%
+    qualification_flag_poisson()
+
+  t10_5_summary <- t10_5_flagged %>%
+    select(GroupID, Numerator, Denominator, Metric, Score, Flag) %>%
+    arrange(desc(abs(Metric))) %>%
+    arrange(match(Flag, c(2, -2, 1, -1, 0)))
+
+
+  t10_5 <- list(
+    "dfTransformed" = t10_5_transformed,
+    "dfAnalyzed" = t10_5_analyzed,
+    "dfFlagged" = t10_5_flagged,
+    "dfSummary" = t10_5_summary
+  )
+
+  # compare results
+  expect_equal(test10_5$lData[names(test10_5$lData) != "dfBounds"], t10_5)
+})
diff --git a/tests/testqualification/qualification/test_qual_T10_6.R b/tests/testqualification/qualification/test_qual_T10_6.R
@@ -0,0 +1,43 @@
+test_that("Query rate assessment can return a correctly assessed data frame for the poisson test grouped by the country variable when given correct input data from clindata and the results should be flagged correctly.", {
+  # gsm analysis
+  dfInput <- gsm::QueryRate_Map_Raw()
+
+  test10_6 <- QueryRate_Assess(
+    dfInput = dfInput,
+    strMethod = "Poisson",
+    strGroup = "Country"
+  )
+
+  # double programming
+  t10_6_input <- dfInput
+
+  t10_6_transformed <- dfInput %>%
+    qualification_transform_counts(countCol = "Count",
+                                   exposureCol = "DataPoint",
+                                   GroupID = "CountryID")
+
+  t10_6_analyzed <- t10_6_transformed %>%
+    qualification_analyze_poisson()
+
+  class(t10_6_analyzed) <- c("tbl_df", "tbl", "data.frame")
+
+
+  t10_6_flagged <- t10_6_analyzed %>%
+    qualification_flag_poisson()
+
+  t10_6_summary <- t10_6_flagged %>%
+    select(GroupID, Numerator, Denominator, Metric, Score, Flag) %>%
+    arrange(desc(abs(Metric)))
E804
 %>%
+    arrange(match(Flag, c(2, -2, 1, -1, 0)))
+
+
+  t10_6 <- list(
+    "dfTransformed" = t10_6_transformed,
+    "dfAnalyzed" = t10_6_analyzed,
+    "dfFlagged" = t10_6_flagged,
+    "dfSummary" = t10_6_summary
+  )
+
+  # compare results
+  expect_equal(test10_6$lData[names(test10_6$lData) != "dfBounds"], t10_6)
+})
diff --git a/tests/testqualification/qualification/test_qual_T10_7.R b/tests/testqualification/qualification/test_qual_T10_7.R
@@ -0,0 +1,42 @@
+test_that("Query rate assessment can return a correctly assessed data frame for the normal approximation test grouped by the site variable when given correct input data from clindata and the results should be flagged correctly using a custom threshold.", {
+  # gsm analysis
+  dfInput <- gsm::QueryRate_Map_Raw()
+
+  test10_7 <- QueryRate_Assess(
+    dfInput = dfInput,
+    strMethod = "NormalApprox",
+    strGroup = "Site",
+    vThreshold = c(-2, -1, 1, 2)
+  )
+
+  # double programming
+  t10_7_input <- dfInput
+
+  t10_7_transformed <- dfInput %>%
+    qualification_transform_counts(countCol = "Count",
+                                   exposureCol = "DataPoint")
+
+  t10_7_analyzed <- t10_7_transformed %>%
+    qualification_analyze_normalapprox(strType = "rate")
+
+  class(t10_7_analyzed) <- c("tbl_df", "tbl", "data.frame")
+
+  t10_7_flagged <- t10_7_analyzed %>%
+    qualification_flag_normalapprox(threshold = c(-2, -1, 1, 2))
+
+  t10_7_summary <- t10_7_flagged %>%
+    select(GroupID, Numerator, Denominator, Metric, Score, Flag) %>%
+    arrange(desc(abs(Metric))) %>%
+    arrange(match(Flag, c(2, -2, 1, -1, 0)))
+
+
+  t10_7 <- list(
+    "dfTransformed" = t10_7_transformed,
+    "dfAnalyzed" = t10_7_analyzed,
+    "dfFlagged" = t10_7_flagged,
+    "dfSummary" = t10_7_summary
+  )
+
+  # compare results
+  expect_equal(test10_7$lData[names(test10_7$lData) != "dfBounds"], t10_7)
+})