tests/testthat/test-plot_precision_recall.R

test_that("plot_precision_recall works", {
  
    data("CnR_H3K27ac")
    data("CnT_H3K27ac")
    data("encode_H3K27ac")
    peakfiles <- list(CnR_H3K27ac=CnR_H3K27ac, CnT_H3K27ac=CnT_H3K27ac)
    reference <- list("encode_H3K27ac" = encode_H3K27ac)

    pr_out <- plot_precision_recall(peakfiles = peakfiles,
                                    reference = reference,
                                    workers = 1)
    testthat::expect_equal(nrow(pr_out$data), 20*length(peakfiles))
    testthat::expect_true(all(c("precision","recall","F1") %in% names(pr_out$data)))
    testthat::expect_true(methods::is(pr_out$precision_recall_plot,"gg"))
    testthat::expect_true(methods::is(pr_out$f1_plot,"gg"))
    
    #### Prediction precision-recall #####
    res <- predict_precision_recall(pr_df = pr_out$data)
    testthat::expect_equal(nrow(res),20*length(peakfiles))
    testthat::expect_equal(round(max(res[predicted_var=="precision",]$predicted)),
                           96)
    testthat::expect_equal(round(max(res[predicted_var=="recall",]$predicted)),
                           53)
    # ggplot(res[predicted_var=="recall",], 
    #        aes(x=predicted, y=input, color=peaklist1)) +
    #     geom_point()
})
neurogenomics/EpiCompare documentation built on April 30, 2024, 3:58 p.m.