Init seed with random value

[koenderks]

See https://github.com/jasp-stats/INTERNAL-jasp/issues/2733

[koenderks]

@lottemensink What do you think about initialising the seeds (in selection, model fairness and number bunching) with random values?

[koenderks]

@JorisGoosen I think this set.seed(1) that is added to all unit tests is a problem in jaspTools. If I make the unit test using:

options <- jaspTools::analysisOptions("auditSelection")
options$id <- "ID"
options$values <- "bookValue"
options$n <- 200
options$explanatoryText <- TRUE
options$tableSample <- TRUE
options$export_sample <- FALSE
options$file <- ""
options$units <- "values"
options$sampling_method <- "interval"
options$startMethod <- "fixedStart"
options$randomize <- FALSE
options$rank <- ""
options$tableDescriptives <- FALSE
options$start <- 1
options$name_indicator <- ""
options$seed <- 1
results <- jaspTools::runAnalysis("auditSelection", "BuildIt_Monetary.csv", options, makeTests = T)

Then I get this test code from jaspResults:

options <- analysisOptions("auditSelection")
options$n <- 200
options$id <- "ID"
options$values <- "bookValue"
options$explanatoryText <- TRUE
options$tableSample <- TRUE
options$export_sample <- FALSE
options$file <- ""
options$units <- "values"
options$sampling_method <- "interval"
options$startMethod <- "fixedStart"
options$randomize <- FALSE
options$rank <- ""
options$tableDescriptives <- FALSE
options$start <- 1
options$name_indicator <- ""
options$seed <- 1
set.seed(1) # COMMENT: NOTE THIS
results <- runAnalysis("auditSelection", "BuildIt_Monetary.csv", options)


test_that("<b>Table 2.</b> Information about Monetary Interval Selection results match", {
	table <- results[["results"]][["selectionContainer"]][["collection"]][["selectionContainer_tableInterval"]][["data"]]
	jaspTools::expect_equal_tables(table,
		list(3500, 1403220.82, 200, 200, "7.5%", "Total", 105247.58, 0, 0,
			 0, 0, "0%", "Top stratum", 0, 3500, 1403220.82, 200, 200, "7.5%",
			 "Bottom stratum", 105247.58))
})

test_that("<b>Table 3.</b> Selected Items results match", {
	table <- results[["results"]][["selectionContainer"]][["collection"]][["selectionContainer_tableSample"]][["data"]]
	jaspTools::expect_equal_tables(table,
		list(82884, 1, 1, 242.61, 46163, 17, 1, 511.62, 57172, 38, 1, 329.3,
			 86073, 55, 1, 874.46, 90160, 73, 1, 205.69, 31740, 89, 1, 735.45,
			 4756, 110, 1, 295.96, 13081, 128, 1, 734.93, 90183, 146, 1,
			 333.28, 76263, 162, 1, 701.99, 96080, 183, 1, 449.07, 96029,
			 201, 1, 465.35, 11890, 218, 1, 296.31, 98624, 235, 1, 340.6,
			 78992, 252, 1, 686.1, 96292, 269, 1, 179.32, 22629, 289, 1,
			 488.41, 11569, 306, 1, 1025.03, 57791, 325, 1, 376, 15197, 344,
			 1, 536.77, 20711, 358, 1, 610.88, 58735, 377, 1, 620.96, 75887,
			 392, 1, 739, 497, 411, 1, 640.18, 24835, 430, 1, 237, 52055,
			 449, 1, 879.91, 26364, 462, 1, 740.25, 71824, 479, 1, 305.14,
			 8847, 497, 1, 397.11, 56756, 514, 1, 348.08, 61005, 536, 1,
			 233.97, 44391, 554, 1, 374.93, 88214, 574, 1, 214.61, 16689,
			 589, 1, 630.99, 27905, 606, 1, 868.59, 51619, 625, 1, 274.09,
			 58370, 641, 1, 182.14, 47485, 660, 1, 773.14, 20237, 679, 1,
			 669.75, 64305, 699, 1, 319.68, 99012, 715, 1, 313.75, 78676,
			 735, 1, 346.28, 4261, 752, 1, 214.07, 47596, 771, 1, 256.13,
			 72660, 795, 1, 366.19, 82711, 813, 1, 288.93, 80057, 828, 1,
			 762.64, 35269, 850, 1, 281.84, 83339, 867, 1, 614.68, 93477,
			 887, 1, 692.36, 98110, 902, 1, 617.6, 73208, 921, 1, 370.28,
			 68039, 939, 1, 684.79, 99419, 956, 1, 668.93, 64369, 972, 1,
			 382.49, 47424, 992, 1, 469.18, 92166, 1010, 1, 566.01, 83189,
			 1031, 1, 199.8, 59197, 1050, 1, 796.94, 37569, 1068, 1, 633.59,
			 65319, 1081, 1, 502.54, 85233, 1098, 1, 686.1, 28178, 1113,
			 1, 367.64, 38019, 1132, 1, 537.14, 37412, 1150, 1, 306.66, 59189,
			 1165, 1, 186.49, 10345, 1182, 1, 294.04, 90284, 1198, 1, 550.7,
			 12827, 1214, 1, 688.28, 5324, 1233, 1, 503.95, 57764, 1249,
			 1, 696.32, 6771, 1267, 1, 351.53, 45561, 1286, 1, 702.76, 60970,
			 1301, 1, 348.48, 28529, 1317, 1, 471.97, 77871, 1335, 1, 812.13,
			 19517, 1354, 1, 340.07, 8070, 1368, 1, 938.12, 83336, 1383,
			 1, 656.34, 55381, 1402, 1, 900.38, 88454, 1421, 1, 856.28, 13214,
			 1437, 1, 994.44, 60630, 1455, 1, 681.91, 83225, 1474, 1, 221.15,
			 81443, 1492, 1, 543.8, 48139, 1505, 1, 718.58, 7652, 1518, 1,
			 846.99, 36945, 1534, 1, 714.02, 4120, 1552, 1, 233.88, 26525,
			 1572, 1, 941.47, 97834, 1590, 1, 294.9, 60760, 1606, 1, 715.14,
			 98301, 1627, 1, 429.07, 6686, 1646, 1, 343.83, 39990, 1668,
			 1, 788, 20736, 1685, 1, 407.24, 83784, 1702, 1, 389.55, 35982,
			 1721, 1, 829.17, 24275, 1738, 1, 721.75, 10054, 1754, 1, 369.34,
			 87258, 1774, 1, 157.68, 6154, 1791, 1, 285.42, 40471, 1808,
			 1, 438.74, 65326, 1822, 1, 484.13, 37044, 1839, 1, 654.61, 78043,
			 1858, 1, 664.6, 78011, 1877, 1, 583.89, 2885, 1895, 1, 285,
			 94914, 1912, 1, 646.29, 27340, 1927, 1, 659.18, 98561, 1942,
			 1, 294.24, 27256, 1958, 1, 314.22, 85588, 1976, 1, 535.51, 2740,
			 1995, 1, 714.23, 5590, 2010, 1, 676.34, 21217, 2026, 1, 172.91,
			 13683, 2041, 1, 381.64, 35189, 2058, 1, 337.29, 29742, 2075,
			 1, 554.85, 89961, 2089, 1, 337.64, 48652, 2103, 1, 497.21, 99274,
			 2118, 1, 816.62, 72040, 2135, 1, 509.04, 25056, 2149, 1, 344.4,
			 16092, 2168, 1, 595.35, 39429, 2183, 1, 338.23, 87666, 2199,
			 1, 544.79, 53398, 2220, 1, 448.74, 60413, 2238, 1, 418.78, 80386,
			 2257, 1, 553.69, 60271, 2275, 1, 502.81, 34881, 2289, 1, 1041.99,
			 93548, 2306, 1, 439.82, 432, 2324, 1, 593.65, 9063, 2340, 1,
			 700.38, 1692, 2354, 1, 607.25, 49157, 2369, 1, 911.21, 23549,
			 2389, 1, 718.86, 65112, 2403, 1, 408.75, 58501, 2420, 1, 386.36,
			 37248, 2435, 1, 1041.44, 53304, 2452, 1, 346.7, 68187, 2470,
			 1, 289.23, 72681, 2486, 1, 651.26, 78346, 2505, 1, 478.69, 37458,
			 2522, 1, 554.27, 64553, 2540, 1, 432.18, 16188, 2558, 1, 649.78,
			 23322, 2575, 1, 378.66, 76001, 2590, 1, 1230.94, 97083, 2605,
			 1, 849.02, 95414, 2618, 1, 337.08, 32760, 2636, 1, 599.3, 39805,
			 2656, 1, 617.63, 35420, 2679, 1, 320.63, 225, 2698, 1, 507.18,
			 46509, 2719, 1, 353.72, 73254, 2736, 1, 343.81, 11585, 2755,
			 1, 157.51, 98790, 2768, 1, 558.36, 10925, 2787, 1, 377.1, 67454,
			 2806, 1, 218.2, 66041, 2824, 1, 684.86, 51035, 2843, 1, 337.96,
			 58155, 2860, 1, 229.59, 76649, 2880, 1, 465.22, 71135, 2900,
			 1, 334.26, 41792, 2920, 1, 202.58, 21214, 2941, 1, 430.11, 81544,
			 2961, 1, 427.78, 99150, 2982, 1, 514.03, 74281, 2998, 1, 215.46,
			 75272, 3014, 1, 510.93, 52121, 3029, 1, 993.57, 13359, 3047,
			 1, 1262.85, 6080, 3064, 1, 722.45, 20335, 3082, 1, 564.01, 78336,
			 3100, 1, 359.43, 31532, 3120, 1, 469, 72035, 3137, 1, 322.92,
			 71832, 3152, 1, 1001.82, 77405, 3169, 1, 907.06, 67180, 3187,
			 1, 254.39, 14754, 3205, 1, 212.81, 90783, 3222, 1, 543.95, 17096,
			 3237, 1, 606.61, 89779, 3255, 1, 295.77, 61391, 3273, 1, 136.1,
			 2111, 3292, 1, 215.07, 79729, 3313, 1, 882.54, 47872, 3330,
			 1, 444.67, 85306, 3349, 1, 674.63, 15, 3368, 1, 475.92, 26052,
			 3383, 1, 681.73, 99612, 3394, 1, 2244.4, 18796, 3409, 1, 414.75,
			 70852, 3428, 1, 606.14, 54606, 3444, 1, 377.4, 53902, 3465,
			 1, 400.3, 89394, 3483, 1, 698.2))
})

test_that("<b>Table 1.</b> Selection Summary results match", {
	table <- results[["results"]][["selectionContainer"]][["collection"]][["selectionContainer_tableSelection"]][["data"]]
	jaspTools::expect_equal_tables(table,
		list(200, "7.5%", 200, 105247.58))
})

So maybe it is a good idea to fix this in jaspTools. Because I just copy and paste the code that jaspTools gives me into the unit tests, and that should probably be the ideal workflow, right? Anyway, I don't think that set.seed(1) actually does something because in the R functions we call set.seed(options$seed).

[JorisGoosen]

You missed the seed in sampling?

This is why its better yo (re)use a component instead of having random integerfields in places :p
Although I guess this code is from before SetSeed being a component anyway

[koenderks]

You missed the seed in sampling?

What do you mean?