tests/testthat/test_build_lm_0.R
In exploratory-io/exploratory_func: R package for Exploratory
context("test build_lm")
test_that("test build_lm summary output ", {
test_df = data.frame(
num1 = seq(20) / 10.0,
num2 = seq(20) - 10,
weight = seq(20) + 100,
category = rep(letters[1:4], 5),
with_NA = rep(c(letters[5:6], NA, NA), 5)
)
model_df <- test_df %>% build_lm(num1 ~ num2 + category + with_NA, weights = weight)
expect_equal(colnames(model_df), c("source.data", ".test_index", "model", ".model_metadata"))
res <- capture.output(summary(model_df$model[[1]]))
expect_lt(length(res), 50) # the output of summary should be less than 50 lines
ret <- model_df %>% tidy_rowwise(model)
ret <- model_df %>% glance_rowwise(model)
ret <- model_df %>% augment_rowwise(model)
ret <- model_df %>% prediction(data = "training") # Test prediction with training data when the training data predictors has NAs.
})
test_that("Test invalid predictors - only one unique value", {
expect_error({
test_df = data.frame(
num1 = runif(20),
num2 = 1,
num3 = 1
)
model_df <- test_df %>% build_lm.fast(num1, num2, num3)
}, "Invalid Predictors: Only one unique value.")
})
test_that("test build_lm with keep.source FALSE ", {
test_df = data.frame(
num1 = seq(20) / 10.0,
num2 = seq(20) - 10,
weight = seq(20) + 100,
category = rep(letters[1:4], 5),
with_NA = rep(c(letters[5:6], NA, NA), 5)
)
trial <- test_df %>% build_lm(num1 ~ num2 + category + with_NA, weights = weight, keep.source = FALSE)
expect_equal(colnames(trial), c(".test_index", "model", ".model_metadata"))
})
test_that("test build_lm with grouped ", {
test_df = data.frame(
num1 = seq(20) / 10.0,
num2 = seq(20) - 10,
group1 = rep(letters[1:4], 5),
group2 = rep(letters[1:2], each = 10)
)
trial <- test_df %>% build_lm(num1 ~ num2, group_cols = c("group1", "group2"))
expect_equal(length(trial[["group2"]]), 8)
expect_equal(length(trial[["group1"]]), 8)
expect_error(build_lm(test_df, num1 ~ num2 + group1, group_cols = c("group1", "group2")), "group1 is a grouping column. Please remove it from variables.")
})
test_that("test build_lm with augment TRUE", {
test_df = data.frame(
num1 = seq(20) / 10.0,
num2 = seq(20) - 10,
group1 = rep(letters[1:4], 5),
group2 = rep(letters[1:2], each = 10)
)
trial <- test_df %>% build_lm(num1 ~ num2, group_cols = c("group1", "group2"), augment = TRUE)
expect_equal(length(trial[["group2"]]), 20)
expect_equal(length(trial[["group1"]]), 20)
})
test_that("test name conflict avoid", {
test_df = data.frame(
num1 = seq(20) / 10.0,
num2 = seq(20) - 11,
estimate = rep(letters[1:4], 5),
model = rep(letters[1:2], each = 10),
model.group = rep(letters[1:2], each = 10)
)
lm_model <- test_df %>%
build_lm(num1 ~ num2, group_cols = c("estimate", "model", "model.group"))
expect_equal(colnames(lm_model), c("estimate.group", "model.group", "model.group1", "source.data", ".test_index", "model", ".model_metadata"))
trial <- suppressWarnings({
lm_model %>%
tidy_rowwise(model)
})
expect_equal(colnames(trial), c("estimate.group", "model.group", "model.group1",
"term", "estimate", "std.error", "statistic", "p.value"))
})
test_that("build_lm with evaluation", {
test_df = data.frame(
num1 = seq(20) / 10.0,
num2 = seq(20) - 11,
group = rep(letters[1:2], each = 10)
)
lm_model <- test_df %>%
build_lm(num1 ~ num2, group_cols = c("group"), test_rate = 0.1)
evaluated <- lm_model %>%
prediction(data = "test")
expect_equal(colnames(evaluated), c("group", "num1", "num2", "predicted_value",
"conf_low", "conf_high",
"standard_error",
"residuals"))
test_eval <- lm_model %>%
prediction(data = "training")
expect_equal(colnames(test_eval), c("group", "num1", "num2",
"predicted_value",
"conf_low", "conf_high",
"standard_error",
"residuals",
"hat", "residual_standard_deviation", "cooks_distance",
"standardised_residuals"
))
})
test_that("prediction with categorical columns", {
test_data <- structure(
list(
CANCELLED = c(0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0),
`Carrier Name` = c("Delta Air Lines", "American Eagle", "American Airlines", "Southwest Airlines", "SkyWest Airlines", "Southwest Airlines", "Southwest Airlines", "Delta Air Lines", "Southwest Airlines", "Atlantic Southeast Airlines", "American Airlines", "Southwest Airlines", "US Airways", "US Airways", "Delta Air Lines", "Atlantic Southeast Airlines", NA, "Atlantic Southeast Airlines", "Delta Air Lines", "Delta Air Lines"),
CARRIER = c("DL", "MQ", "AA", "DL", "MQ", "AA", "DL", "DL", "MQ", "AA", "AA", "WN", "US", "US", "DL", "EV", "9E", "EV", "DL", "DL"),
DISTANCE = c(1587, 173, 646, 187, 273, 1062, 583, 240, 1123, 851, 852, 862, 361, 507, 1020, 1092, 342, 489, 1184, 545)), row.names = c(NA, -20L),
class = c("tbl_df", "tbl", "data.frame"), .Names = c("CANCELLED", "Carrier Name", "CARRIER", "DISTANCE"))
# duplicate rows to make some predictable data
# otherwise, the number of rows of the result of prediction becomes 0
test_data <- dplyr::bind_rows(test_data, test_data)
model_data <- build_lm(test_data, CANCELLED ~ `Carrier Name` + CARRIER + DISTANCE, test_rate = 0.6)
ret <- prediction(model_data, data = "test")
expect_true(nrow(ret) > 0)
expect_true(all(c("CANCELLED",
"Carrier Name", "CARRIER", "DISTANCE",
"predicted_value",
"conf_low", "conf_high",
"standard_error",
"residuals") %in% colnames(ret)))
grouped <- test_data %>%
dplyr::group_by(CARRIER)
expect_error(build_lm(grouped, CANCELLED ~ CARRIER), "CARRIER is a grouping column. Please remove it from variables.")
})
test_that("prediction with target column name with space by build_lm.fast", {
test_data <- structure(
list(
`CANCELLED:X` = c(0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0),
`logical col` = c(TRUE, FALSE, TRUE, FALSE, FALSE, TRUE, FALSE, FALSE, NA, TRUE, FALSE, NA, FALSE, TRUE, FALSE, TRUE, FALSE, FALSE, FALSE, FALSE),
`Carrier-Name` = c("Delta Air Lines", "American Eagle", "American Airlines", "Southwest Airlines", "SkyWest Airlines", "Southwest Airlines", "Southwest Airlines", "Delta Air Lines", "Southwest Airlines", "Atlantic Southeast Airlines", "American Airlines", "Southwest Airlines", "US Airways", "US Airways", "Delta Air Lines", "Atlantic Southeast Airlines", NA, "Atlantic Southeast Airlines", "Delta Air Lines", "Delta Air Lines"),
CARRIER = c("DL", "MQ", "AA", "DL", "MQ", "AA", "DL", "DL", "MQ", "AA", "AA", "WN", "US", "US", "DL", "EV", "9E", "EV", "DL", "DL"),
# testing filtering of Inf, -Inf, NA here.
DISTANCE = c(Inf, -Inf, NA, 187, 273, 1062, 583, 240, 1123, 851, 852, 862, 361, 507, 1020, 1092, 342, 489, 1184, 545)), row.names = c(NA, -20L),
class = c("tbl_df", "tbl", "data.frame"), .Names = c("CANCELLED:X", "logical col", "Carrier-Name", "CARRIER", "DISTANCE"))
# duplicate rows to make some predictable data
# otherwise, the number of rows of the result of prediction becomes 0
test_data <- dplyr::bind_rows(test_data, test_data)
test_data <- test_data %>% mutate(`Carrier-Name`= factor(`Carrier-Name`, ordered=TRUE)) # test handling of ordered factor
model_data <- build_lm.fast(test_data, `CANCELLED:X`, `logical col`, `Carrier-Name`, DISTANCE, predictor_n = 3, with_marginal_effects=TRUE, with_marginal_effects_confint=TRUE)
ret <- model_data %>% glance_rowwise(model)
# TODO: the returned coefficients does not show all input variables.
# most likely due to too few rows. look into it and add check for the values in the returned df.
ret <- model_data %>% tidy_rowwise(model)
# Verify that base levels are not NA for `logical col` (testing space in the name) and Carrier-Name (testing - in the name) columns.
ret2 <- ret %>% dplyr::filter(stringr::str_detect(term,"(logical col|Carrier-Name)")) %>% dplyr::summarize(na_count=sum(is.na(base.level)))
expect_equal(ret2$na_count, 0)
ret <- model_data %>% augment_rowwise(model)
expect_true(nrow(ret) > 0)
expect_equal(colnames(ret), c("CANCELLED:X", "Carrier-Name","DISTANCE","logical col", ".fitted",".se.fit",".resid",".hat",".sigma",".cooksd",".std.resid"))
})
test_that("prediction with glm family (binomial) and link (probit) with target column name with space by build_lm.fast", {
test_data <- structure(
list(
`CANCELLED X` = c(0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0),
`logical col` = c(TRUE, FALSE, TRUE, FALSE, FALSE, TRUE, FALSE, FALSE, NA, TRUE, FALSE, NA, FALSE, TRUE, FALSE, TRUE, FALSE, FALSE, FALSE, FALSE),
`Carrier Name` = c("Delta Air Lines", "American Eagle", "American Airlines", "Southwest Airlines", "SkyWest Airlines", "Southwest Airlines", "Southwest Airlines", "Delta Air Lines", "Southwest Airlines", "Atlantic Southeast Airlines", "American Airlines", "Southwest Airlines", "US Airways", "US Airways", "Delta Air Lines", "Atlantic Southeast Airlines", NA, "Atlantic Southeast Airlines", "Delta Air Lines", "Delta Air Lines"),
CARRIER = c("DL", "MQ", "AA", "DL", "MQ", "AA", "DL", "DL", "MQ", "AA", "AA", "WN", "US", "US", "DL", "EV", "9E", "EV", "DL", "DL"),
# testing filtering of Inf, -Inf, NA here.
DISTANCE = c(Inf, -Inf, NA, 187, 273, 1062, 583, 240, 1123, 851, 852, 862, 361, 507, 1020, 1092, 342, 489, 1184, 545)), row.names = c(NA, -20L),
class = c("tbl_df", "tbl", "data.frame"), .Names = c("CANCELLED X", "logical col", "Carrier Name", "CARRIER", "DISTANCE"))
# duplicate rows to make some predictable data
# otherwise, the number of rows of the result of prediction becomes 0
test_data <- dplyr::bind_rows(test_data, test_data)
test_data <- test_data %>% mutate(CARRIER = factor(CARRIER, ordered=TRUE)) # test handling of ordered factor
# should run without error. TODO: verify resulting values.
model_data <- build_lm.fast(test_data, `CANCELLED X`, `logical col`, `Carrier Name`, CARRIER, DISTANCE, predictor_n = 3, model_type = "glm", family = "poisson", link = "log")
ret <- model_data %>% glance_rowwise(model)
ret <- model_data %>% tidy_rowwise(model)
ret <- model_data %>% augment_rowwise(model)
# should run without error. TODO: verify resulting values.
model_data <- build_lm.fast(test_data, `CANCELLED X`, `logical col`, `Carrier Name`, CARRIER, DISTANCE, predictor_n = 3, model_type = "glm", family = "gaussian", link = "identity")
ret <- model_data %>% glance_rowwise(model)
ret <- model_data %>% tidy_rowwise(model)
ret <- model_data %>% augment_rowwise(model)
# should run without error. TODO: verify resulting values.
model_data <- build_lm.fast(test_data, `CANCELLED X`, `logical col`, `Carrier Name`, CARRIER, DISTANCE, predictor_n = 3, model_type = "glm", family = "binomial", link = "probit")
ret <- model_data %>% glance_rowwise(model)
# TODO: the returned coefficients does not show all input variables.
# most likely due to too few rows. look into it and add check for the values in the returned df.
ret <- model_data %>% tidy_rowwise(model)
ret <- model_data %>% augment_rowwise(model)
expect_true(nrow(ret) > 0)
expect_true(all(c("CANCELLED X", "logical col", "Carrier Name","CARRIER","DISTANCE",".fitted",".se.fit",".resid",".hat",".sigma",".cooksd",".std.resid", "predicted_response", "predicted_label")
%in% colnames(ret)))
})
test_that("prediction with glm family (negativebinomial) with target column name with space by build_lm.fast", {
test_data <- structure(
list(
`CANCELLED X` = c(0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0),
`logical col` = c(TRUE, FALSE, TRUE, FALSE, FALSE, TRUE, FALSE, FALSE, NA, TRUE, FALSE, NA, FALSE, TRUE, FALSE, TRUE, FALSE, FALSE, FALSE, FALSE),
`Carrier Name` = c("Delta Air Lines", "American Eagle", "American Airlines", "Southwest Airlines", "SkyWest Airlines", "Southwest Airlines", "Southwest Airlines", "Delta Air Lines", "Southwest Airlines", "Atlantic Southeast Airlines", "American Airlines", "Southwest Airlines", "US Airways", "US Airways", "Delta Air Lines", "Atlantic Southeast Airlines", NA, "Atlantic Southeast Airlines", "Delta Air Lines", "Delta Air Lines"),
CARRIER = c("DL", "MQ", "AA", "DL", "MQ", "AA", "DL", "DL", "MQ", "AA", "AA", "WN", "US", "US", "DL", "EV", "9E", "EV", "DL", "DL"),
# testing filtering of Inf, -Inf, NA here.
DISTANCE = c(Inf, -Inf, NA, 187, 273, 1062, 583, 240, 1123, 851, 852, 862, 361, 507, 1020, 1092, 342, 489, 1184, 545)), row.names = c(NA, -20L),
class = c("tbl_df", "tbl", "data.frame"), .Names = c("CANCELLED X", "logical col", "Carrier Name", "CARRIER", "DISTANCE"))
model_data <- build_lm.fast(test_data,
`CANCELLED X`,
`logical col`,
`Carrier Name`,
CARRIER, DISTANCE,
predictor_n = 3,
model_type = "glm",
link = "log",
family="negativebinomial")
ret <- model_data %>% glance_rowwise(model, pretty.name=TRUE)
expect_equal(colnames(ret),
c("P Value", "Rows", "Log Likelihood", "AIC", "BIC", "Residual Deviance", "Residual DF", "Null Deviance", "Null Model DF",
"Theta", "SE Theta", "Max VIF"))
ret <- model_data %>% tidy_rowwise(model)
expect_colnames <- c("term", "estimate", "std.error", "statistic", "p.value",
"conf.low", "conf.high", "base.level")
# when model has NA value in coefficients, broom::tidy(model) return note column
expect_true(identical(colnames(ret), expect_colnames) ||
identical(colnames(ret), c(expect_colnames, "note")))
ret <- model_data %>% augment_rowwise(model)
expect_true(all(colnames(ret) %in%
c("CANCELLED X",
"logical col", "Carrier Name", "CARRIER", "DISTANCE",
".fitted", ".resid", ".std.resid", ".hat", ".sigma", ".cooksd", "predicted_response")))
})
if (Sys.info()["sysname"] != "Windows") {
test_that("prediction with target column name with Japanese by build_lm.fast", {
test_data <- structure(
list(
`キャンセル X` = c(0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0),
`論理 col` = c(TRUE, FALSE, TRUE, FALSE, FALSE, TRUE, FALSE, FALSE, NA, TRUE, FALSE, NA, FALSE, TRUE, FALSE, TRUE, FALSE, FALSE, FALSE, FALSE),
`航空会社 Name` = c("Delta Air Lines", "American Eagle", "American Airlines", "Southwest Airlines", "SkyWest Airlines", "Southwest Airlines", "Southwest Airlines", "Delta Air Lines", "Southwest Airlines", "Atlantic Southeast Airlines", "American Airlines", "Southwest Airlines", "US Airways", "US Airways", "Delta Air Lines", "Atlantic Southeast Airlines", NA, "Atlantic Southeast Airlines", "Delta Air Lines", "Delta Air Lines"),
CARRIER = c("DL", "MQ", "AA", "DL", "MQ", "AA", "DL", "DL", "MQ", "AA", "AA", "WN", "US", "US", "DL", "EV", "9E", "EV", "DL", "DL"),
# testing filtering of Inf, -Inf, NA here.
DISTANCE = c(Inf, -Inf, NA, 187, 273, 1062, 583, 240, 1123, 851, 852, 862, 361, 507, 1020, 1092, 342, 489, 1184, 545)), row.names = c(NA, -20L),
class = c("tbl_df", "tbl", "data.frame"), .Names = c("キャンセル X", "論理 col", "航空会社 Name", "CARRIER", "DISTANCE"))
# duplicate rows to make some predictable data
# otherwise, the number of rows of the result of prediction becomes 0
test_data <- dplyr::bind_rows(test_data, test_data)
test_data <- test_data %>% mutate(CARRIER = factor(CARRIER, ordered=TRUE)) # test handling of ordered factor
model_data <- build_lm.fast(test_data, `キャンセル X`, `論理 col`, `航空会社 Name`, CARRIER, DISTANCE, predictor_n = 3, with_marginal_effects=TRUE, with_marginal_effects_confint=TRUE)
ret <- model_data %>% glance_rowwise(model)
# TODO: the returned coefficients does not show all input variables.
# most likely due to too few rows. look into it and add check for the values in the returned df.
ret <- model_data %>% tidy_rowwise(model)
ret <- model_data %>% augment_rowwise(model)
expect_true(nrow(ret) > 0)
expect_true(all(colnames(ret) %in% c("キャンセル X", "論理 col", "航空会社 Name","CARRIER","DISTANCE",".fitted",".se.fit",".resid",".hat",".sigma",".cooksd",".std.resid")))
})
}
test_that("prediction with glm model with SMOTE by build_lm.fast", {
test_data <- structure(
list(
`CANCELLED X` = c(0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0),
`Carrier Name` = c("Delta Air Lines", "American Eagle", "American Airlines", "Southwest Airlines", "SkyWest Airlines", "Southwest Airlines", "Southwest Airlines", "Delta Air Lines", "Southwest Airlines", "Atlantic Southeast Airlines", "American Airlines", "Southwest Airlines", "US Airways", "US Airways", "Delta Air Lines", "Atlantic Southeast Airlines", NA, "Atlantic Southeast Airlines", "Delta Air Lines", "Delta Air Lines"),
CARRIER = factor(c(NA, "MQ", "AA", "DL", "MQ", "AA", "DL", "DL", "MQ", "AA", "AA", "WN", "US", "US", "DL", "EV", "9E", "EV", "DL", "DL")), # test with factor with NA
# testing filtering of Inf, -Inf, NA here.
DISTANCE = c(Inf, -Inf, NA, 187, 273, 1062, 583, 240, 1123, 851, 852, 862, 361, 507, 1020, 1092, 342, 489, 1184, 545)), row.names = c(NA, -20L),
class = c("tbl_df", "tbl", "data.frame"), .Names = c("CANCELLED X", "Carrier Name", "CARRIER", "DISTANCE"))
# duplicate rows to make some predictable data
# otherwise, the number of rows of the result of prediction becomes 0
test_data <- dplyr::bind_rows(test_data, test_data)
test_data <- test_data %>% mutate(CARRIER = factor(CARRIER, ordered=TRUE)) # test handling of ordered factor
model_data <- build_lm.fast(test_data, `CANCELLED X`, `Carrier Name`, CARRIER, DISTANCE, model_type = "glm", smote=FALSE, with_marginal_effects=TRUE, with_marginal_effects_confint=TRUE)
ret <- model_data %>% lm_partial_dependence()
ret <- model_data %>% glance_rowwise(model, pretty.name=TRUE)
ret <- model_data %>% tidy_rowwise(model)
ret <- model_data %>% augment_rowwise(model)
model_data <- build_lm.fast(test_data, `CANCELLED X`, `Carrier Name`, CARRIER, DISTANCE, model_type = "glm", smote=TRUE, with_marginal_effects=TRUE, with_marginal_effects_confint=TRUE)
ret <- model_data %>% glance_rowwise(model, pretty.name=TRUE)
ret <- model_data %>% tidy_rowwise(model)
ret <- model_data %>% augment_rowwise(model)
# test for perfect multicollinearity case.
reduced_test_data <- test_data %>% tail(3)
model_data <- build_lm.fast(reduced_test_data, `CANCELLED X`, `Carrier Name`, CARRIER, DISTANCE, model_type = "glm", smote=FALSE, with_marginal_effects=TRUE, with_marginal_effects_confint=TRUE)
ret <- model_data %>% glance_rowwise(model, pretty.name=TRUE)
ret <- model_data %>% tidy_rowwise(model, pretty.name=TRUE)
ret <- model_data %>% augment_rowwise(model, pretty.name=TRUE)
expect_true(nrow(ret) > 0)
})
test_that("test GLM (Negative Binomial) summary output", {
test_df <- data.frame(
num1 = seq(20),
num2 = seq(20) - 10,
num3 = seq(20) / 10.0
)
ret <- test_df %>% build_lm.fast(num1, num2, num3,
model_type="glm",
family="negativebinomial")
model_ret <- ret %>% glance_rowwise(model)
expect_equal(colnames(model_ret),
c("null.deviance", "df.null", "logLik",
"AIC", "BIC", "deviance", "df.residual",
"p.value", "n", "theta", "SE.theta"))
model_ret_pretty <- ret %>% glance_rowwise(model, pretty.name=TRUE)
expect_equal(colnames(model_ret_pretty),
c("P Value", "Rows", "Log Likelihood", "AIC",
"BIC", "Residual Deviance", "Residual DF", "Null Deviance",
"Null Model DF", "Theta", "SE Theta"))
})
test_that("test GLM (Negative Binomial) with group columns", {
test_data <- structure(
list(
`CANCELLED X` = c(0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0),
`Carrier Name` = c("Delta Air Lines", "American Eagle", "American Airlines", "Southwest Airlines", "SkyWest Airlines", "Southwest Airlines", "Southwest Airlines", "Delta Air Lines", "Southwest Airlines", "Atlantic Southeast Airlines", "American Airlines", "Southwest Airlines", "US Airways", "US Airways", "Delta Air Lines", "Atlantic Southeast Airlines", NA, "Atlantic Southeast Airlines", "Delta Air Lines", "Delta Air Lines"),
CARRIER = factor(c(NA, "MQ", "AA", "DL", "MQ", "AA", "DL", "DL", "MQ", "AA", "AA", "WN", "US", "US", "DL", "EV", "9E", "EV", "DL", "DL")), # test with factor with NA
# testing filtering of Inf, -Inf, NA here.
DISTANCE = c(Inf, -Inf, NA, 187, 273, 1062, 583, 240, 1123, 851, 852, 862, 361, 507, 1020, 1092, 342, 489, 1184, 545)), row.names = c(NA, -20L),
class = c("tbl_df", "tbl", "data.frame"), .Names = c("CANCELLED X", "Carrier Name", "CARRIER", "DISTANCE"))
# duplicate rows to make some predictable data
# otherwise, the number of rows of the result of prediction becomes 0
test_data <- dplyr::bind_rows(test_data, test_data)
test_data <- test_data %>% mutate(CARRIER = factor(CARRIER, ordered=TRUE)) # test handling of ordered factor
ret <- test_data %>% dplyr::group_by(CARRIER) %>%
build_lm.fast(`CANCELLED X`,
DISTANCE,
predictor_n = 3,
model_type = "glm",
family = "negativebinomial")
expect_equal(length(ret[["CARRIER"]]), 8)
model_ret <- ret %>% glance_rowwise(model)
expect_equal(colnames(model_ret), # Position of Note columns is adjusted on Exploratory-side
c("CARRIER", "Note", "null.deviance", "df.null", "logLik",
"AIC", "BIC", "deviance", "df.residual",
"p.value", "n", "theta", "SE.theta"))
model_ret_pretty <- ret %>% glance_rowwise(model, pretty.name=TRUE)
expect_equal(colnames(model_ret_pretty), # Position of Note columns is adjusted on Exploratory-side
c("CARRIER", "Note", "P Value", "Rows", "Log Likelihood", "AIC",
"BIC", "Residual Deviance", "Residual DF", "Null Deviance",
"Null Model DF", "Theta", "SE Theta"))
})
test_that("test GLM (Negative Binomial) with group columns with weight", {
test_data <- structure(
list(
`CANCELLED X` = c(0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0),
`Carrier Name` = c("Delta Air Lines", "American Eagle", "American Airlines", "Southwest Airlines", "SkyWest Airlines", "Southwest Airlines", "Southwest Airlines", "Delta Air Lines", "Southwest Airlines", "Atlantic Southeast Airlines", "American Airlines", "Southwest Airlines", "US Airways", "US Airways", "Delta Air Lines", "Atlantic Southeast Airlines", NA, "Atlantic Southeast Airlines", "Delta Air Lines", "Delta Air Lines"),
CARRIER = factor(c(NA, "MQ", "AA", "DL", "MQ", "AA", "DL", "DL", "MQ", "AA", "AA", "WN", "US", "US", "DL", "EV", "9E", "EV", "DL", "DL")), # test with factor with NA
# testing filtering of Inf, -Inf, NA here.
DISTANCE = c(Inf, -Inf, NA, 187, 273, 1062, 583, 240, 1123, 851, 852, 862, 361, 507, 1020, 1092, 342, 489, 1184, 545)), row.names = c(NA, -20L),
class = c("tbl_df", "tbl", "data.frame"), .Names = c("CANCELLED X", "Carrier Name", "CARRIER", "DISTANCE"))
# duplicate rows to make some predictable data
# otherwise, the number of rows of the result of prediction becomes 0
test_data <- dplyr::bind_rows(test_data, test_data)
test_data <- test_data %>% mutate(CARRIER = factor(CARRIER, ordered=TRUE)) # test handling of ordered factor
set.seed(0);
test_data <- test_data %>% mutate(Weight=0.3*sin(1:n())+1)
ret <- test_data %>% dplyr::group_by(CARRIER) %>%
build_lm.fast(`CANCELLED X`,
DISTANCE,
weight=Weight,
predictor_n = 3,
model_type = "glm",
family = "negativebinomial")
expect_equal(length(ret[["CARRIER"]]), 8)
model_ret <- ret %>% glance_rowwise(model)
# Check the numbers so that we can detect any change in broom or stats in the future.
expect_equal((ret %>% tidy_rowwise(model))$estimate, c(-174.5406548, 0.1474161), tolerance=1e-4)
expect_equal(colnames(model_ret), # Position of Note columns is adjusted on Exploratory-side
c("CARRIER", "Note", "null.deviance", "df.null", "logLik",
"AIC", "BIC", "deviance", "df.residual",
"p.value", "n", "theta", "SE.theta"))
model_ret_pretty <- ret %>% glance_rowwise(model, pretty.name=TRUE)
expect_equal(colnames(model_ret_pretty), # Position of Note columns is adjusted on Exploratory-side
c("CARRIER", "Note", "P Value", "Rows", "Log Likelihood", "AIC",
"BIC", "Residual Deviance", "Residual DF", "Null Deviance",
"Null Model DF", "Theta", "SE Theta"))
})
exploratory-io/exploratory_func documentation built on April 23, 2024, 9:15 p.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
context("test build_lm")
test_that("test build_lm summary output ", {
test_df = data.frame(
num1 = seq(20) / 10.0,
num2 = seq(20) - 10,
weight = seq(20) + 100,
category = rep(letters[1:4], 5),
with_NA = rep(c(letters[5:6], NA, NA), 5)
)
model_df <- test_df %>% build_lm(num1 ~ num2 + category + with_NA, weights = weight)
expect_equal(colnames(model_df), c("source.data", ".test_index", "model", ".model_metadata"))
res <- capture.output(summary(model_df$model[[1]]))
expect_lt(length(res), 50) # the output of summary should be less than 50 lines
ret <- model_df %>% tidy_rowwise(model)
ret <- model_df %>% glance_rowwise(model)
ret <- model_df %>% augment_rowwise(model)
ret <- model_df %>% prediction(data = "training") # Test prediction with training data when the training data predictors has NAs.
})
test_that("Test invalid predictors - only one unique value", {
expect_error({
test_df = data.frame(
num1 = runif(20),
num2 = 1,
num3 = 1
)
model_df <- test_df %>% build_lm.fast(num1, num2, num3)
}, "Invalid Predictors: Only one unique value.")
})
test_that("test build_lm with keep.source FALSE ", {
test_df = data.frame(
num1 = seq(20) / 10.0,
num2 = seq(20) - 10,
weight = seq(20) + 100,
category = rep(letters[1:4], 5),
with_NA = rep(c(letters[5:6], NA, NA), 5)
)
trial <- test_df %>% build_lm(num1 ~ num2 + category + with_NA, weights = weight, keep.source = FALSE)
expect_equal(colnames(trial), c(".test_index", "model", ".model_metadata"))
})
test_that("test build_lm with grouped ", {
test_df = data.frame(
num1 = seq(20) / 10.0,
num2 = seq(20) - 10,
group1 = rep(letters[1:4], 5),
group2 = rep(letters[1:2], each = 10)
)
trial <- test_df %>% build_lm(num1 ~ num2, group_cols = c("group1", "group2"))
expect_equal(length(trial[["group2"]]), 8)
expect_equal(length(trial[["group1"]]), 8)
expect_error(build_lm(test_df, num1 ~ num2 + group1, group_cols = c("group1", "group2")), "group1 is a grouping column. Please remove it from variables.")
})
test_that("test build_lm with augment TRUE", {
test_df = data.frame(
num1 = seq(20) / 10.0,
num2 = seq(20) - 10,
group1 = rep(letters[1:4], 5),
group2 = rep(letters[1:2], each = 10)
)
trial <- test_df %>% build_lm(num1 ~ num2, group_cols = c("group1", "group2"), augment = TRUE)
expect_equal(length(trial[["group2"]]), 20)
expect_equal(length(trial[["group1"]]), 20)
})
test_that("test name conflict avoid", {
test_df = data.frame(
num1 = seq(20) / 10.0,
num2 = seq(20) - 11,
estimate = rep(letters[1:4], 5),
model = rep(letters[1:2], each = 10),
model.group = rep(letters[1:2], each = 10)
)
lm_model <- test_df %>%
build_lm(num1 ~ num2, group_cols = c("estimate", "model", "model.group"))
expect_equal(colnames(lm_model), c("estimate.group", "model.group", "model.group1", "source.data", ".test_index", "model", ".model_metadata"))
trial <- suppressWarnings({
lm_model %>%
tidy_rowwise(model)
})
expect_equal(colnames(trial), c("estimate.group", "model.group", "model.group1",
"term", "estimate", "std.error", "statistic", "p.value"))
})
test_that("build_lm with evaluation", {
test_df = data.frame(
num1 = seq(20) / 10.0,
num2 = seq(20) - 11,
group = rep(letters[1:2], each = 10)
)
lm_model <- test_df %>%
build_lm(num1 ~ num2, group_cols = c("group"), test_rate = 0.1)
evaluated <- lm_model %>%
prediction(data = "test")
expect_equal(colnames(evaluated), c("group", "num1", "num2", "predicted_value",
"conf_low", "conf_high",
"standard_error",
"residuals"))
test_eval <- lm_model %>%
prediction(data = "training")
expect_equal(colnames(test_eval), c("group", "num1", "num2",
"predicted_value",
"conf_low", "conf_high",
"standard_error",
"residuals",
"hat", "residual_standard_deviation", "cooks_distance",
"standardised_residuals"
))
})
test_that("prediction with categorical columns", {
test_data <- structure(
list(
CANCELLED = c(0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0),
`Carrier Name` = c("Delta Air Lines", "American Eagle", "American Airlines", "Southwest Airlines", "SkyWest Airlines", "Southwest Airlines", "Southwest Airlines", "Delta Air Lines", "Southwest Airlines", "Atlantic Southeast Airlines", "American Airlines", "Southwest Airlines", "US Airways", "US Airways", "Delta Air Lines", "Atlantic Southeast Airlines", NA, "Atlantic Southeast Airlines", "Delta Air Lines", "Delta Air Lines"),
CARRIER = c("DL", "MQ", "AA", "DL", "MQ", "AA", "DL", "DL", "MQ", "AA", "AA", "WN", "US", "US", "DL", "EV", "9E", "EV", "DL", "DL"),
DISTANCE = c(1587, 173, 646, 187, 273, 1062, 583, 240, 1123, 851, 852, 862, 361, 507, 1020, 1092, 342, 489, 1184, 545)), row.names = c(NA, -20L),
class = c("tbl_df", "tbl", "data.frame"), .Names = c("CANCELLED", "Carrier Name", "CARRIER", "DISTANCE"))
# duplicate rows to make some predictable data
# otherwise, the number of rows of the result of prediction becomes 0
test_data <- dplyr::bind_rows(test_data, test_data)
model_data <- build_lm(test_data, CANCELLED ~ `Carrier Name` + CARRIER + DISTANCE, test_rate = 0.6)
ret <- prediction(model_data, data = "test")
expect_true(nrow(ret) > 0)
expect_true(all(c("CANCELLED",
"Carrier Name", "CARRIER", "DISTANCE",
"predicted_value",
"conf_low", "conf_high",
"standard_error",
"residuals") %in% colnames(ret)))
grouped <- test_data %>%
dplyr::group_by(CARRIER)
expect_error(build_lm(grouped, CANCELLED ~ CARRIER), "CARRIER is a grouping column. Please remove it from variables.")
})
test_that("prediction with target column name with space by build_lm.fast", {
test_data <- structure(
list(
`CANCELLED:X` = c(0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0),
`logical col` = c(TRUE, FALSE, TRUE, FALSE, FALSE, TRUE, FALSE, FALSE, NA, TRUE, FALSE, NA, FALSE, TRUE, FALSE, TRUE, FALSE, FALSE, FALSE, FALSE),
`Carrier-Name` = c("Delta Air Lines", "American Eagle", "American Airlines", "Southwest Airlines", "SkyWest Airlines", "Southwest Airlines", "Southwest Airlines", "Delta Air Lines", "Southwest Airlines", "Atlantic Southeast Airlines", "American Airlines", "Southwest Airlines", "US Airways", "US Airways", "Delta Air Lines", "Atlantic Southeast Airlines", NA, "Atlantic Southeast Airlines", "Delta Air Lines", "Delta Air Lines"),
CARRIER = c("DL", "MQ", "AA", "DL", "MQ", "AA", "DL", "DL", "MQ", "AA", "AA", "WN", "US", "US", "DL", "EV", "9E", "EV", "DL", "DL"),
# testing filtering of Inf, -Inf, NA here.
DISTANCE = c(Inf, -Inf, NA, 187, 273, 1062, 583, 240, 1123, 851, 852, 862, 361, 507, 1020, 1092, 342, 489, 1184, 545)), row.names = c(NA, -20L),
class = c("tbl_df", "tbl", "data.frame"), .Names = c("CANCELLED:X", "logical col", "Carrier-Name", "CARRIER", "DISTANCE"))
# duplicate rows to make some predictable data
# otherwise, the number of rows of the result of prediction becomes 0
test_data <- dplyr::bind_rows(test_data, test_data)
test_data <- test_data %>% mutate(`Carrier-Name`= factor(`Carrier-Name`, ordered=TRUE)) # test handling of ordered factor
model_data <- build_lm.fast(test_data, `CANCELLED:X`, `logical col`, `Carrier-Name`, DISTANCE, predictor_n = 3, with_marginal_effects=TRUE, with_marginal_effects_confint=TRUE)
ret <- model_data %>% glance_rowwise(model)
# TODO: the returned coefficients does not show all input variables.
# most likely due to too few rows. look into it and add check for the values in the returned df.
ret <- model_data %>% tidy_rowwise(model)
# Verify that base levels are not NA for `logical col` (testing space in the name) and Carrier-Name (testing - in the name) columns.
ret2 <- ret %>% dplyr::filter(stringr::str_detect(term,"(logical col|Carrier-Name)")) %>% dplyr::summarize(na_count=sum(is.na(base.level)))
expect_equal(ret2$na_count, 0)
ret <- model_data %>% augment_rowwise(model)
expect_true(nrow(ret) > 0)
expect_equal(colnames(ret), c("CANCELLED:X", "Carrier-Name","DISTANCE","logical col", ".fitted",".se.fit",".resid",".hat",".sigma",".cooksd",".std.resid"))
})
test_that("prediction with glm family (binomial) and link (probit) with target column name with space by build_lm.fast", {
test_data <- structure(
list(
`CANCELLED X` = c(0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0),
`logical col` = c(TRUE, FALSE, TRUE, FALSE, FALSE, TRUE, FALSE, FALSE, NA, TRUE, FALSE, NA, FALSE, TRUE, FALSE, TRUE, FALSE, FALSE, FALSE, FALSE),
`Carrier Name` = c("Delta Air Lines", "American Eagle", "American Airlines", "Southwest Airlines", "SkyWest Airlines", "Southwest Airlines", "Southwest Airlines", "Delta Air Lines", "Southwest Airlines", "Atlantic Southeast Airlines", "American Airlines", "Southwest Airlines", "US Airways", "US Airways", "Delta Air Lines", "Atlantic Southeast Airlines", NA, "Atlantic Southeast Airlines", "Delta Air Lines", "Delta Air Lines"),
CARRIER = c("DL", "MQ", "AA", "DL", "MQ", "AA", "DL", "DL", "MQ", "AA", "AA", "WN", "US", "US", "DL", "EV", "9E", "EV", "DL", "DL"),
# testing filtering of Inf, -Inf, NA here.
DISTANCE = c(Inf, -Inf, NA, 187, 273, 1062, 583, 240, 1123, 851, 852, 862, 361, 507, 1020, 1092, 342, 489, 1184, 545)), row.names = c(NA, -20L),
class = c("tbl_df", "tbl", "data.frame"), .Names = c("CANCELLED X", "logical col", "Carrier Name", "CARRIER", "DISTANCE"))
# duplicate rows to make some predictable data
# otherwise, the number of rows of the result of prediction becomes 0
test_data <- dplyr::bind_rows(test_data, test_data)
test_data <- test_data %>% mutate(CARRIER = factor(CARRIER, ordered=TRUE)) # test handling of ordered factor
# should run without error. TODO: verify resulting values.
model_data <- build_lm.fast(test_data, `CANCELLED X`, `logical col`, `Carrier Name`, CARRIER, DISTANCE, predictor_n = 3, model_type = "glm", family = "poisson", link = "log")
ret <- model_data %>% glance_rowwise(model)
ret <- model_data %>% tidy_rowwise(model)
ret <- model_data %>% augment_rowwise(model)
# should run without error. TODO: verify resulting values.
model_data <- build_lm.fast(test_data, `CANCELLED X`, `logical col`, `Carrier Name`, CARRIER, DISTANCE, predictor_n = 3, model_type = "glm", family = "gaussian", link = "identity")
ret <- model_data %>% glance_rowwise(model)
ret <- model_data %>% tidy_rowwise(model)
ret <- model_data %>% augment_rowwise(model)
# should run without error. TODO: verify resulting values.
model_data <- build_lm.fast(test_data, `CANCELLED X`, `logical col`, `Carrier Name`, CARRIER, DISTANCE, predictor_n = 3, model_type = "glm", family = "binomial", link = "probit")
ret <- model_data %>% glance_rowwise(model)
# TODO: the returned coefficients does not show all input variables.
# most likely due to too few rows. look into it and add check for the values in the returned df.
ret <- model_data %>% tidy_rowwise(model)
ret <- model_data %>% augment_rowwise(model)
expect_true(nrow(ret) > 0)
expect_true(all(c("CANCELLED X", "logical col", "Carrier Name","CARRIER","DISTANCE",".fitted",".se.fit",".resid",".hat",".sigma",".cooksd",".std.resid", "predicted_response", "predicted_label")
%in% colnames(ret)))
})
test_that("prediction with glm family (negativebinomial) with target column name with space by build_lm.fast", {
test_data <- structure(
list(
`CANCELLED X` = c(0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0),
`logical col` = c(TRUE, FALSE, TRUE, FALSE, FALSE, TRUE, FALSE, FALSE, NA, TRUE, FALSE, NA, FALSE, TRUE, FALSE, TRUE, FALSE, FALSE, FALSE, FALSE),
`Carrier Name` = c("Delta Air Lines", "American Eagle", "American Airlines", "Southwest Airlines", "SkyWest Airlines", "Southwest Airlines", "Southwest Airlines", "Delta Air Lines", "Southwest Airlines", "Atlantic Southeast Airlines", "American Airlines", "Southwest Airlines", "US Airways", "US Airways", "Delta Air Lines", "Atlantic Southeast Airlines", NA, "Atlantic Southeast Airlines", "Delta Air Lines", "Delta Air Lines"),
CARRIER = c("DL", "MQ", "AA", "DL", "MQ", "AA", "DL", "DL", "MQ", "AA", "AA", "WN", "US", "US", "DL", "EV", "9E", "EV", "DL", "DL"),
# testing filtering of Inf, -Inf, NA here.
DISTANCE = c(Inf, -Inf, NA, 187, 273, 1062, 583, 240, 1123, 851, 852, 862, 361, 507, 1020, 1092, 342, 489, 1184, 545)), row.names = c(NA, -20L),
class = c("tbl_df", "tbl", "data.frame"), .Names = c("CANCELLED X", "logical col", "Carrier Name", "CARRIER", "DISTANCE"))
model_data <- build_lm.fast(test_data,
`CANCELLED X`,
`logical col`,
`Carrier Name`,
CARRIER, DISTANCE,
predictor_n = 3,
model_type = "glm",
link = "log",
family="negativebinomial")
ret <- model_data %>% glance_rowwise(model, pretty.name=TRUE)
expect_equal(colnames(ret),
c("P Value", "Rows", "Log Likelihood", "AIC", "BIC", "Residual Deviance", "Residual DF", "Null Deviance", "Null Model DF",
"Theta", "SE Theta", "Max VIF"))
ret <- model_data %>% tidy_rowwise(model)
expect_colnames <- c("term", "estimate", "std.error", "statistic", "p.value",
"conf.low", "conf.high", "base.level")
# when model has NA value in coefficients, broom::tidy(model) return note column
expect_true(identical(colnames(ret), expect_colnames) ||
identical(colnames(ret), c(expect_colnames, "note")))
ret <- model_data %>% augment_rowwise(model)
expect_true(all(colnames(ret) %in%
c("CANCELLED X",
"logical col", "Carrier Name", "CARRIER", "DISTANCE",
".fitted", ".resid", ".std.resid", ".hat", ".sigma", ".cooksd", "predicted_response")))
})
if (Sys.info()["sysname"] != "Windows") {
test_that("prediction with target column name with Japanese by build_lm.fast", {
test_data <- structure(
list(
`キャンセル X` = c(0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0),
`論理 col` = c(TRUE, FALSE, TRUE, FALSE, FALSE, TRUE, FALSE, FALSE, NA, TRUE, FALSE, NA, FALSE, TRUE, FALSE, TRUE, FALSE, FALSE, FALSE, FALSE),
`航空会社 Name` = c("Delta Air Lines", "American Eagle", "American Airlines", "Southwest Airlines", "SkyWest Airlines", "Southwest Airlines", "Southwest Airlines", "Delta Air Lines", "Southwest Airlines", "Atlantic Southeast Airlines", "American Airlines", "Southwest Airlines", "US Airways", "US Airways", "Delta Air Lines", "Atlantic Southeast Airlines", NA, "Atlantic Southeast Airlines", "Delta Air Lines", "Delta Air Lines"),
CARRIER = c("DL", "MQ", "AA", "DL", "MQ", "AA", "DL", "DL", "MQ", "AA", "AA", "WN", "US", "US", "DL", "EV", "9E", "EV", "DL", "DL"),
# testing filtering of Inf, -Inf, NA here.
DISTANCE = c(Inf, -Inf, NA, 187, 273, 1062, 583, 240, 1123, 851, 852, 862, 361, 507, 1020, 1092, 342, 489, 1184, 545)), row.names = c(NA, -20L),
class = c("tbl_df", "tbl", "data.frame"), .Names = c("キャンセル X", "論理 col", "航空会社 Name", "CARRIER", "DISTANCE"))
# duplicate rows to make some predictable data
# otherwise, the number of rows of the result of prediction becomes 0
test_data <- dplyr::bind_rows(test_data, test_data)
test_data <- test_data %>% mutate(CARRIER = factor(CARRIER, ordered=TRUE)) # test handling of ordered factor
model_data <- build_lm.fast(test_data, `キャンセル X`, `論理 col`, `航空会社 Name`, CARRIER, DISTANCE, predictor_n = 3, with_marginal_effects=TRUE, with_marginal_effects_confint=TRUE)
ret <- model_data %>% glance_rowwise(model)
# TODO: the returned coefficients does not show all input variables.
# most likely due to too few rows. look into it and add check for the values in the returned df.
ret <- model_data %>% tidy_rowwise(model)
ret <- model_data %>% augment_rowwise(model)
expect_true(nrow(ret) > 0)
expect_true(all(colnames(ret) %in% c("キャンセル X", "論理 col", "航空会社 Name","CARRIER","DISTANCE",".fitted",".se.fit",".resid",".hat",".sigma",".cooksd",".std.resid")))
})
}
test_that("prediction with glm model with SMOTE by build_lm.fast", {
test_data <- structure(
list(
`CANCELLED X` = c(0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0),
`Carrier Name` = c("Delta Air Lines", "American Eagle", "American Airlines", "Southwest Airlines", "SkyWest Airlines", "Southwest Airlines", "Southwest Airlines", "Delta Air Lines", "Southwest Airlines", "Atlantic Southeast Airlines", "American Airlines", "Southwest Airlines", "US Airways", "US Airways", "Delta Air Lines", "Atlantic Southeast Airlines", NA, "Atlantic Southeast Airlines", "Delta Air Lines", "Delta Air Lines"),
CARRIER = factor(c(NA, "MQ", "AA", "DL", "MQ", "AA", "DL", "DL", "MQ", "AA", "AA", "WN", "US", "US", "DL", "EV", "9E", "EV", "DL", "DL")), # test with factor with NA
# testing filtering of Inf, -Inf, NA here.
DISTANCE = c(Inf, -Inf, NA, 187, 273, 1062, 583, 240, 1123, 851, 852, 862, 361, 507, 1020, 1092, 342, 489, 1184, 545)), row.names = c(NA, -20L),
class = c("tbl_df", "tbl", "data.frame"), .Names = c("CANCELLED X", "Carrier Name", "CARRIER", "DISTANCE"))
# duplicate rows to make some predictable data
# otherwise, the number of rows of the result of prediction becomes 0
test_data <- dplyr::bind_rows(test_data, test_data)
test_data <- test_data %>% mutate(CARRIER = factor(CARRIER, ordered=TRUE)) # test handling of ordered factor
model_data <- build_lm.fast(test_data, `CANCELLED X`, `Carrier Name`, CARRIER, DISTANCE, model_type = "glm", smote=FALSE, with_marginal_effects=TRUE, with_marginal_effects_confint=TRUE)
ret <- model_data %>% lm_partial_dependence()
ret <- model_data %>% glance_rowwise(model, pretty.name=TRUE)
ret <- model_data %>% tidy_rowwise(model)
ret <- model_data %>% augment_rowwise(model)
model_data <- build_lm.fast(test_data, `CANCELLED X`, `Carrier Name`, CARRIER, DISTANCE, model_type = "glm", smote=TRUE, with_marginal_effects=TRUE, with_marginal_effects_confint=TRUE)
ret <- model_data %>% glance_rowwise(model, pretty.name=TRUE)
ret <- model_data %>% tidy_rowwise(model)
ret <- model_data %>% augment_rowwise(model)
# test for perfect multicollinearity case.
reduced_test_data <- test_data %>% tail(3)
model_data <- build_lm.fast(reduced_test_data, `CANCELLED X`, `Carrier Name`, CARRIER, DISTANCE, model_type = "glm", smote=FALSE, with_marginal_effects=TRUE, with_marginal_effects_confint=TRUE)
ret <- model_data %>% glance_rowwise(model, pretty.name=TRUE)
ret <- model_data %>% tidy_rowwise(model, pretty.name=TRUE)
ret <- model_data %>% augment_rowwise(model, pretty.name=TRUE)
expect_true(nrow(ret) > 0)
})
test_that("test GLM (Negative Binomial) summary output", {
test_df <- data.frame(
num1 = seq(20),
num2 = seq(20) - 10,
num3 = seq(20) / 10.0
)
ret <- test_df %>% build_lm.fast(num1, num2, num3,
model_type="glm",
family="negativebinomial")
model_ret <- ret %>% glance_rowwise(model)
expect_equal(colnames(model_ret),
c("null.deviance", "df.null", "logLik",
"AIC", "BIC", "deviance", "df.residual",
"p.value", "n", "theta", "SE.theta"))
model_ret_pretty <- ret %>% glance_rowwise(model, pretty.name=TRUE)
expect_equal(colnames(model_ret_pretty),
c("P Value", "Rows", "Log Likelihood", "AIC",
"BIC", "Residual Deviance", "Residual DF", "Null Deviance",
"Null Model DF", "Theta", "SE Theta"))
})
test_that("test GLM (Negative Binomial) with group columns", {
test_data <- structure(
list(
`CANCELLED X` = c(0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0),
`Carrier Name` = c("Delta Air Lines", "American Eagle", "American Airlines", "Southwest Airlines", "SkyWest Airlines", "Southwest Airlines", "Southwest Airlines", "Delta Air Lines", "Southwest Airlines", "Atlantic Southeast Airlines", "American Airlines", "Southwest Airlines", "US Airways", "US Airways", "Delta Air Lines", "Atlantic Southeast Airlines", NA, "Atlantic Southeast Airlines", "Delta Air Lines", "Delta Air Lines"),
CARRIER = factor(c(NA, "MQ", "AA", "DL", "MQ", "AA", "DL", "DL", "MQ", "AA", "AA", "WN", "US", "US", "DL", "EV", "9E", "EV", "DL", "DL")), # test with factor with NA
# testing filtering of Inf, -Inf, NA here.
DISTANCE = c(Inf, -Inf, NA, 187, 273, 1062, 583, 240, 1123, 851, 852, 862, 361, 507, 1020, 1092, 342, 489, 1184, 545)), row.names = c(NA, -20L),
class = c("tbl_df", "tbl", "data.frame"), .Names = c("CANCELLED X", "Carrier Name", "CARRIER", "DISTANCE"))
# duplicate rows to make some predictable data
# otherwise, the number of rows of the result of prediction becomes 0
test_data <- dplyr::bind_rows(test_data, test_data)
test_data <- test_data %>% mutate(CARRIER = factor(CARRIER, ordered=TRUE)) # test handling of ordered factor
ret <- test_data %>% dplyr::group_by(CARRIER) %>%
build_lm.fast(`CANCELLED X`,
DISTANCE,
predictor_n = 3,
model_type = "glm",
family = "negativebinomial")
expect_equal(length(ret[["CARRIER"]]), 8)
model_ret <- ret %>% glance_rowwise(model)
expect_equal(colnames(model_ret), # Position of Note columns is adjusted on Exploratory-side
c("CARRIER", "Note", "null.deviance", "df.null", "logLik",
"AIC", "BIC", "deviance", "df.residual",
"p.value", "n", "theta", "SE.theta"))
model_ret_pretty <- ret %>% glance_rowwise(model, pretty.name=TRUE)
expect_equal(colnames(model_ret_pretty), # Position of Note columns is adjusted on Exploratory-side
c("CARRIER", "Note", "P Value", "Rows", "Log Likelihood", "AIC",
"BIC", "Residual Deviance", "Residual DF", "Null Deviance",
"Null Model DF", "Theta", "SE Theta"))
})
test_that("test GLM (Negative Binomial) with group columns with weight", {
test_data <- structure(
list(
`CANCELLED X` = c(0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0),
`Carrier Name` = c("Delta Air Lines", "American Eagle", "American Airlines", "Southwest Airlines", "SkyWest Airlines", "Southwest Airlines", "Southwest Airlines", "Delta Air Lines", "Southwest Airlines", "Atlantic Southeast Airlines", "American Airlines", "Southwest Airlines", "US Airways", "US Airways", "Delta Air Lines", "Atlantic Southeast Airlines", NA, "Atlantic Southeast Airlines", "Delta Air Lines", "Delta Air Lines"),
CARRIER = factor(c(NA, "MQ", "AA", "DL", "MQ", "AA", "DL", "DL", "MQ", "AA", "AA", "WN", "US", "US", "DL", "EV", "9E", "EV", "DL", "DL")), # test with factor with NA
# testing filtering of Inf, -Inf, NA here.
DISTANCE = c(Inf, -Inf, NA, 187, 273, 1062, 583, 240, 1123, 851, 852, 862, 361, 507, 1020, 1092, 342, 489, 1184, 545)), row.names = c(NA, -20L),
class = c("tbl_df", "tbl", "data.frame"), .Names = c("CANCELLED X", "Carrier Name", "CARRIER", "DISTANCE"))
# duplicate rows to make some predictable data
# otherwise, the number of rows of the result of prediction becomes 0
test_data <- dplyr::bind_rows(test_data, test_data)
test_data <- test_data %>% mutate(CARRIER = factor(CARRIER, ordered=TRUE)) # test handling of ordered factor
set.seed(0);
test_data <- test_data %>% mutate(Weight=0.3*sin(1:n())+1)
ret <- test_data %>% dplyr::group_by(CARRIER) %>%
build_lm.fast(`CANCELLED X`,
DISTANCE,
weight=Weight,
predictor_n = 3,
model_type = "glm",
family = "negativebinomial")
expect_equal(length(ret[["CARRIER"]]), 8)
model_ret <- ret %>% glance_rowwise(model)
# Check the numbers so that we can detect any change in broom or stats in the future.
expect_equal((ret %>% tidy_rowwise(model))$estimate, c(-174.5406548, 0.1474161), tolerance=1e-4)
expect_equal(colnames(model_ret), # Position of Note columns is adjusted on Exploratory-side
c("CARRIER", "Note", "null.deviance", "df.null", "logLik",
"AIC", "BIC", "deviance", "df.residual",
"p.value", "n", "theta", "SE.theta"))
model_ret_pretty <- ret %>% glance_rowwise(model, pretty.name=TRUE)
expect_equal(colnames(model_ret_pretty), # Position of Note columns is adjusted on Exploratory-side
c("CARRIER", "Note", "P Value", "Rows", "Log Likelihood", "AIC",
"BIC", "Residual Deviance", "Residual DF", "Null Deviance",
"Null Model DF", "Theta", "SE Theta"))
})
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.