tests/testthat/test_cascade.R

In desmarais-lab/NetworkInference: Inferring Latent Diffusion Networks

library(NetworkInference)

set.seed(2552)
context("Test if cascade data structure and related methods work.")

n_casc <- 100
n_nodes <- 100

# Long format
test_that("Long format", {
    dat <- simulate_rnd_cascades(n_casc, n_nodes)
    casc <- as_cascade_long(data = dat, cascade_node_name = "node_name", 
                            event_time = "event_time", 
                            cascade_id = "cascade_id")
    df <- as.data.frame(casc)
    df <- df[order(df$cascade_id, df$node_name), ]
    dat$cascade_id <- as.character(dat$cascade_id)
    dat <- dat[order(dat$cascade_id, dat$node_name), ]
    rownames(dat) <- c(1:nrow(dat))
    rownames(df) <- c(1:nrow(df))
    expect_equal(dat, df)    
})

# Wide format
test_that("Wide format", {
    dat <- simulate_rnd_cascades(n_casc, n_nodes)
    casc <- as_cascade_long(data = dat, cascade_node_name = "node_name", 
                            event_time = "event_time", 
                            cascade_id = "cascade_id")
    m <- as.matrix(casc)
    cascade <- as_cascade_wide(m)
    expect_equal(casc, cascade)
})

# Subsetting
test_that("Subsetting works", {
    data("cascades") 
    reduced_cascade <- subset_cascade_time(cascades, 10, 20, drop=TRUE)
    expect_equal(length(reduced_cascade), length(cascades))
    expect_equal(length(reduced_cascade$cascade_nodes), length(reduced_cascade$cascade_times))
    expect_equal(min(do.call(c, reduced_cascade$cascade_times)), 10.1, tolerance = 0.1)
    expect_equal(max(do.call(c, reduced_cascade$cascade_times)), 19.8, tolerance = 0.1)
    reduced_cascade <- subset_cascade_time(cascades, 10, 20, drop=FALSE)
    expect_equal(length(reduced_cascade), length(cascades))
    expect_equal(length(reduced_cascade$cascade_nodes), length(reduced_cascade$cascade_times))
    expect_equal(length(reduced_cascade$cascade_nodes), length(cascades$cascade_nodes))
    expect_equal(length(reduced_cascade$cascade_times), length(cascades$cascade_times))
    expect_equal(min(do.call(c, reduced_cascade$cascade_times)), 10.1, tolerance = 0.1)
    expect_equal(max(do.call(c, reduced_cascade$cascade_times)), 19.8, tolerance = 0.1)
    expect_warning(as.data.frame(reduced_cascade))
})