Convergence diagnostics (R-hat and ESS) for a SNP-Slice run
Source:R/convergence.R
convergence_diagnostics.RdReports rank-normalized split-R-hat and effective sample size (bulk and tail), pooled across all chains, for permutation-invariant summaries of the run. One row is returned per parameter for the whole run rather than one per chain, so R-hat reflects between-chain agreement and ESS reflects the total effective sample count. A single-chain run still yields a (split-)R-hat.
Usage
convergence_diagnostics(
results,
pars = c("logpost", "n_strains", "kstar", "ktrunc"),
additional_burnin = 0
)Arguments
- results
A
snp_slice_resultsobject (withstore_mcmc = TRUE).- pars
Character vector of parameters to include. One of
"logpost","n_strains","kstar","ktrunc", which are all scalar per sample, or"coi"(per-host complexity of infection), which is a vector that expands tocoi[1]..coi[N]).- additional_burnin
Number of additional stored samples to discard from the start of each retained chain before assembling the array.
Value
A data frame with one row per variable and columns variable,
mean, median, sd, q5, q95, rhat,
ess_bulk, ess_tail.
Details
Constant parameters (e.g. kstar or ktrunc are frequently constant
across retained samples) have zero variance, for which posterior returns
NaN R-hat and ESS. That means "this parameter did not move", not that
the diagnostic failed. To diagnose the A/D matrices, use
"coi", which is invariant to strain relabeling; element-wise R-hat / ESS
on A/D is intentionally not provided (see
as_draws_snp_slice).
Examples
result <- load_example_results()
convergence_diagnostics(result)
#> variable mean median sd q5 q95 rhat
#> 1 logpost -75055.92244 -75010.39 122.577982 -75254.45 -74902.59 2.187591
#> 2 n_strains 51.06133 52.00 2.157725 48.00 53.00 7.593054
#> 3 kstar 113.00000 113.00 0.000000 113.00 113.00 NA
#> 4 ktrunc 134.00000 134.00 0.000000 134.00 134.00 NA
#> ess_bulk ess_tail
#> 1 4.008605 41.678197
#> 2 3.322269 3.522944
#> 3 NA NA
#> 4 NA NA
# Per-host complexity of infection
convergence_diagnostics(result, pars = "coi")
#> variable mean median sd q5 q95 rhat ess_bulk ess_tail
#> 1 coi[1] 1.000000 1 0.00000000 1 1 NA NA NA
#> 2 coi[2] 1.000000 1 0.00000000 1 1 NA NA NA
#> 3 coi[3] 1.000000 1 0.00000000 1 1 NA NA NA
#> 4 coi[4] 1.000000 1 0.00000000 1 1 NA NA NA
#> 5 coi[5] 1.000000 1 0.00000000 1 1 NA NA NA
#> 6 coi[6] 1.000000 1 0.00000000 1 1 NA NA NA
#> 7 coi[7] 1.000000 1 0.00000000 1 1 NA NA NA
#> 8 coi[8] 1.000000 1 0.00000000 1 1 NA NA NA
#> 9 coi[9] 1.000000 1 0.00000000 1 1 NA NA NA
#> 10 coi[10] 1.000000 1 0.00000000 1 1 NA NA NA
#> 11 coi[11] 1.000000 1 0.00000000 1 1 NA NA NA
#> 12 coi[12] 1.000000 1 0.00000000 1 1 NA NA NA
#> 13 coi[13] 1.000000 1 0.00000000 1 1 NA NA NA
#> 14 coi[14] 1.000000 1 0.00000000 1 1 NA NA NA
#> 15 coi[15] 2.000000 2 0.00000000 2 2 NA NA NA
#> 16 coi[16] 1.000000 1 0.00000000 1 1 NA NA NA
#> 17 coi[17] 1.000000 1 0.00000000 1 1 NA NA NA
#> 18 coi[18] 1.000000 1 0.00000000 1 1 NA NA NA
#> 19 coi[19] 1.000000 1 0.00000000 1 1 NA NA NA
#> 20 coi[20] 2.000000 2 0.00000000 2 2 NA NA NA
#> 21 coi[21] 1.000000 1 0.00000000 1 1 NA NA NA
#> 22 coi[22] 1.000000 1 0.00000000 1 1 NA NA NA
#> 23 coi[23] 1.000000 1 0.00000000 1 1 NA NA NA
#> 24 coi[24] 1.000000 1 0.00000000 1 1 NA NA NA
#> 25 coi[25] 1.000000 1 0.00000000 1 1 NA NA NA
#> 26 coi[26] 1.000000 1 0.00000000 1 1 NA NA NA
#> 27 coi[27] 1.000000 1 0.00000000 1 1 NA NA NA
#> 28 coi[28] 1.000000 1 0.00000000 1 1 NA NA NA
#> 29 coi[29] 1.000000 1 0.00000000 1 1 NA NA NA
#> 30 coi[30] 1.000000 1 0.00000000 1 1 NA NA NA
#> 31 coi[31] 1.000000 1 0.00000000 1 1 NA NA NA
#> 32 coi[32] 3.333333 4 0.94343821 2 4 Inf 3.125000 NA
#> 33 coi[33] 1.000000 1 0.00000000 1 1 NA NA NA
#> 34 coi[34] 1.000000 1 0.00000000 1 1 NA NA NA
#> 35 coi[35] 1.000000 1 0.00000000 1 1 NA NA NA
#> 36 coi[36] 1.000000 1 0.00000000 1 1 NA NA NA
#> 37 coi[37] 2.000000 2 0.00000000 2 2 NA NA NA
#> 38 coi[38] 1.000000 1 0.00000000 1 1 NA NA NA
#> 39 coi[39] 1.000000 1 0.00000000 1 1 NA NA NA
#> 40 coi[40] 1.000000 1 0.00000000 1 1 NA NA NA
#> 41 coi[41] 1.000000 1 0.00000000 1 1 NA NA NA
#> 42 coi[42] 1.000000 1 0.00000000 1 1 NA NA NA
#> 43 coi[43] 3.000000 3 0.00000000 3 3 NA NA NA
#> 44 coi[44] 1.000000 1 0.00000000 1 1 NA NA NA
#> 45 coi[45] 1.000000 1 0.00000000 1 1 NA NA NA
#> 46 coi[46] 1.000000 1 0.00000000 1 1 NA NA NA
#> 47 coi[47] 1.000000 1 0.00000000 1 1 NA NA NA
#> 48 coi[48] 1.000000 1 0.00000000 1 1 NA NA NA
#> 49 coi[49] 1.000000 1 0.00000000 1 1 NA NA NA
#> 50 coi[50] 3.333333 3 0.47171911 3 4 Inf 3.125000 NA
#> 51 coi[51] 1.000000 1 0.00000000 1 1 NA NA NA
#> 52 coi[52] 2.000000 2 0.00000000 2 2 NA NA NA
#> 53 coi[53] 1.000000 1 0.00000000 1 1 NA NA NA
#> 54 coi[54] 1.000000 1 0.00000000 1 1 NA NA NA
#> 55 coi[55] 2.000000 2 0.00000000 2 2 NA NA NA
#> 56 coi[56] 1.000000 1 0.00000000 1 1 NA NA NA
#> 57 coi[57] 1.000000 1 0.00000000 1 1 NA NA NA
#> 58 coi[58] 1.000000 1 0.00000000 1 1 NA NA NA
#> 59 coi[59] 1.000000 1 0.00000000 1 1 NA NA NA
#> 60 coi[60] 1.000000 1 0.00000000 1 1 NA NA NA
#> 61 coi[61] 1.000000 1 0.00000000 1 1 NA NA NA
#> 62 coi[62] 1.000000 1 0.00000000 1 1 NA NA NA
#> 63 coi[63] 1.000000 1 0.00000000 1 1 NA NA NA
#> 64 coi[64] 1.000000 1 0.00000000 1 1 NA NA NA
#> 65 coi[65] 1.000000 1 0.00000000 1 1 NA NA NA
#> 66 coi[66] 1.000000 1 0.00000000 1 1 NA NA NA
#> 67 coi[67] 1.000000 1 0.00000000 1 1 NA NA NA
#> 68 coi[68] 1.000000 1 0.00000000 1 1 NA NA NA
#> 69 coi[69] 1.000000 1 0.00000000 1 1 NA NA NA
#> 70 coi[70] 1.000000 1 0.00000000 1 1 NA NA NA
#> 71 coi[71] 1.000000 1 0.00000000 1 1 NA NA NA
#> 72 coi[72] 1.000000 1 0.00000000 1 1 NA NA NA
#> 73 coi[73] 1.000000 1 0.00000000 1 1 NA NA NA
#> 74 coi[74] 2.000000 2 0.00000000 2 2 NA NA NA
#> 75 coi[75] 1.000000 1 0.00000000 1 1 NA NA NA
#> 76 coi[76] 1.000000 1 0.00000000 1 1 NA NA NA
#> 77 coi[77] 2.666667 2 0.94343821 2 4 Inf 3.125000 NA
#> 78 coi[78] 3.338667 4 0.95000433 2 4 11.592553 3.298667 3.125000
#> 79 coi[79] 1.000000 1 0.00000000 1 1 NA NA NA
#> 80 coi[80] 1.000000 1 0.00000000 1 1 NA NA NA
#> 81 coi[81] 1.000000 1 0.00000000 1 1 NA NA NA
#> 82 coi[82] 1.000000 1 0.00000000 1 1 NA NA NA
#> 83 coi[83] 1.000000 1 0.00000000 1 1 NA NA NA
#> 84 coi[84] 2.000000 2 0.00000000 2 2 NA NA NA
#> 85 coi[85] 1.000000 1 0.00000000 1 1 NA NA NA
#> 86 coi[86] 1.000000 1 0.00000000 1 1 NA NA NA
#> 87 coi[87] 4.000000 4 0.00000000 4 4 NA NA NA
#> 88 coi[88] 1.000000 1 0.00000000 1 1 NA NA NA
#> 89 coi[89] 4.000000 4 0.00000000 4 4 NA NA NA
#> 90 coi[90] 1.000000 1 0.00000000 1 1 NA NA NA
#> 91 coi[91] 1.000000 1 0.00000000 1 1 NA NA NA
#> 92 coi[92] 1.000000 1 0.00000000 1 1 NA NA NA
#> 93 coi[93] 2.000000 2 0.00000000 2 2 NA NA NA
#> 94 coi[94] 1.000000 1 0.00000000 1 1 NA NA NA
#> 95 coi[95] 3.620000 4 1.21555341 2 5 4.519544 3.310766 NA
#> 96 coi[96] 2.666667 2 0.94343821 2 4 Inf 3.125000 NA
#> 97 coi[97] 2.666667 2 0.94343821 2 4 Inf 3.125000 NA
#> 98 coi[98] 1.000000 1 0.00000000 1 1 NA NA NA
#> 99 coi[99] 2.000000 2 0.00000000 2 2 NA NA NA
#> 100 coi[100] 1.000000 1 0.00000000 1 1 NA NA NA
#> 101 coi[101] 1.000000 1 0.00000000 1 1 NA NA NA
#> 102 coi[102] 1.000000 1 0.00000000 1 1 NA NA NA
#> 103 coi[103] 4.000000 4 0.00000000 4 4 NA NA NA
#> 104 coi[104] 3.666667 4 0.47171911 3 4 Inf 3.125000 NA
#> 105 coi[105] 1.000000 1 0.00000000 1 1 NA NA NA
#> 106 coi[106] 1.000000 1 0.00000000 1 1 NA NA NA
#> 107 coi[107] 1.000000 1 0.00000000 1 1 NA NA NA
#> 108 coi[108] 1.000000 1 0.00000000 1 1 NA NA NA
#> 109 coi[109] 1.000000 1 0.00000000 1 1 NA NA NA
#> 110 coi[110] 1.000000 1 0.00000000 1 1 NA NA NA
#> 111 coi[111] 3.882667 4 0.73231227 3 5 4.316849 3.307996 NA
#> 112 coi[112] 3.666667 3 0.94343821 3 5 Inf 3.125000 NA
#> 113 coi[113] 1.000000 1 0.00000000 1 1 NA NA NA
#> 114 coi[114] 1.000000 1 0.00000000 1 1 NA NA NA
#> 115 coi[115] 3.000000 3 0.00000000 3 3 NA NA NA
#> 116 coi[116] 1.000000 1 0.00000000 1 1 NA NA NA
#> 117 coi[117] 2.000000 2 0.00000000 2 2 NA NA NA
#> 118 coi[118] 1.000000 1 0.00000000 1 1 NA NA NA
#> 119 coi[119] 2.000000 2 0.00000000 2 2 NA NA NA
#> 120 coi[120] 7.174667 7 0.39033495 7 8 1.273060 9.084968 8.979333
#> 121 coi[121] 1.000000 1 0.00000000 1 1 NA NA NA
#> 122 coi[122] 2.000000 2 0.00000000 2 2 NA NA NA
#> 123 coi[123] 2.000000 2 0.00000000 2 2 NA NA NA
#> 124 coi[124] 1.000000 1 0.00000000 1 1 NA NA NA
#> 125 coi[125] 1.000000 1 0.00000000 1 1 NA NA NA
#> 126 coi[126] 1.000000 1 0.00000000 1 1 NA NA NA
#> 127 coi[127] 1.000000 1 0.00000000 1 1 NA NA NA
#> 128 coi[128] 5.504000 6 0.50031766 5 6 1.654580 5.241860 NA
#> 129 coi[129] 2.000000 2 0.00000000 2 2 NA NA NA
#> 130 coi[130] 1.000000 1 0.00000000 1 1 NA NA NA
#> 131 coi[131] 2.000000 2 0.00000000 2 2 NA NA NA
#> 132 coi[132] 3.333333 4 0.94343821 2 4 Inf 3.125000 NA
#> 133 coi[133] 2.000000 2 0.00000000 2 2 NA NA NA
#> 134 coi[134] 3.666667 3 0.94343821 3 5 Inf 3.125000 NA
#> 135 coi[135] 5.333333 6 0.94343821 4 6 Inf 3.125000 NA
#> 136 coi[136] 4.076000 4 0.27016288 4 5 1.048169 40.873581 41.788425
#> 137 coi[137] 1.000000 1 0.00000000 1 1 NA NA NA
#> 138 coi[138] 6.002667 6 0.05160529 6 6 1.004057 762.293482 762.293482
#> 139 coi[139] 4.000000 4 0.00000000 4 4 NA NA NA
#> 140 coi[140] 7.000000 7 0.00000000 7 7 NA NA NA
#> 141 coi[141] 3.998667 3 1.41374082 3 6 36.816044 3.127322 NA
#> 142 coi[142] 5.838667 7 1.34724100 4 7 3.074668 3.545036 NA
#> 143 coi[143] 3.000000 3 0.00000000 3 3 NA NA NA
#> 144 coi[144] 2.000000 2 0.00000000 2 2 NA NA NA
#> 145 coi[145] 4.000000 4 0.00000000 4 4 NA NA NA
#> 146 coi[146] 1.000000 1 0.00000000 1 1 NA NA NA
#> 147 coi[147] 4.001333 4 0.03651484 4 4 1.000000 762.294282 762.294282
#> 148 coi[148] 5.000000 5 0.81704146 4 6 Inf 3.125000 NA
#> 149 coi[149] 5.676000 6 0.46831231 5 6 5.382374 3.241102 NA
#> 150 coi[150] 1.000000 1 0.00000000 1 1 NA NA NA
#> 151 coi[151] 2.650667 3 0.47707759 2 3 4.150149 3.322690 NA
#> 152 coi[152] 1.000000 1 0.00000000 1 1 NA NA NA
#> 153 coi[153] 5.553333 6 0.49747918 5 6 1.361907 7.300401 NA
#> 154 coi[154] 2.000000 2 0.00000000 2 2 NA NA NA
#> 155 coi[155] 5.000000 5 0.00000000 5 5 NA NA NA
#> 156 coi[156] 2.000000 2 0.00000000 2 2 NA NA NA
#> 157 coi[157] 2.333333 2 0.47171911 2 3 Inf 3.125000 NA
#> 158 coi[158] 5.400000 5 0.49294082 5 6 1.528101 5.665731 5.607288
#> 159 coi[159] 4.670667 4 0.95258391 4 6 6.997293 3.190911 3.125000
#> 160 coi[160] 2.020000 2 0.14009343 2 2 1.061731 51.522678 51.522678
#> 161 coi[161] 3.666667 3 0.94343821 3 5 Inf 3.125000 NA
#> 162 coi[162] 5.806667 6 0.59701825 5 7 1.090452 28.982693 NA
#> 163 coi[163] 6.988000 7 0.45767742 6 8 1.249045 9.180430 NA
#> 164 coi[164] 4.332000 4 0.47406996 4 5 14.149205 3.183924 NA
#> 165 coi[165] 4.000000 4 0.81704146 3 5 Inf 3.125000 NA
#> 166 coi[166] 8.390667 8 0.52765277 8 9 1.064434 85.114262 60.192041
#> 167 coi[167] 4.000000 4 0.00000000 4 4 NA NA NA
#> 168 coi[168] 5.666667 6 0.47171911 5 6 Inf 3.125000 NA
#> 169 coi[169] 3.994667 4 0.21299149 4 4 1.053467 436.138362 252.276117
#> 170 coi[170] 10.282667 11 0.95846636 9 11 3.336735 3.876070 3.125000
#> 171 coi[171] 6.990667 7 0.81943569 6 8 3.444510 3.423619 3.874497
#> 172 coi[172] 6.698667 7 0.54183916 6 7 0.999457 590.325016 224.487002
#> 173 coi[173] 9.189333 10 0.90622728 8 10 3.172876 3.508958 NA
#> 174 coi[174] 5.336000 5 0.47827014 5 6 7.118896 3.188446 3.125000
#> 175 coi[175] 4.218667 5 0.97508253 3 5 1.330181 8.181374 NA
#> 176 coi[176] 9.444000 9 0.64452920 9 11 2.512105 3.749016 NA
#> 177 coi[177] 6.717333 7 1.42271984 5 8 1.640780 5.393832 3.977886
#> 178 coi[178] 8.001333 8 0.81622292 7 9 24.474797 3.130256 NA
#> 179 coi[179] 6.030667 6 0.17252800 6 6 1.035689 158.342515 158.342515
#> 180 coi[180] 6.014667 6 0.16679924 6 6 1.016674 144.537083 146.153999
#> 181 coi[181] 4.666667 5 0.47171911 4 5 Inf 3.125000 NA
#> 182 coi[182] 3.000000 3 0.00000000 3 3 NA NA NA
#> 183 coi[183] 7.000000 7 0.81704146 6 8 Inf 3.125000 NA
#> 184 coi[184] 7.666667 8 0.47171911 7 8 Inf 3.125000 NA
#> 185 coi[185] 6.849333 7 0.35796239 6 7 1.007034 746.787621 NA
#> 186 coi[186] 7.170667 7 0.56624529 6 8 2.063114 4.189021 NA
#> 187 coi[187] 6.640000 7 1.23521969 5 8 7.233266 3.248123 NA
#> 188 coi[188] 5.666667 5 0.94343821 5 7 Inf 3.125000 NA
#> 189 coi[189] 6.445333 7 1.77253135 4 8 4.285125 3.324352 NA
#> 190 coi[190] 5.373333 5 0.48401231 5 6 1.008974 652.173023 NA
#> 191 coi[191] 4.558667 4 0.72612756 4 6 1.372938 7.346099 4.479165
#> 192 coi[192] 7.326667 7 1.24499890 6 9 6.966828 3.191522 NA
#> 193 coi[193] 5.333333 5 0.47171911 5 6 Inf 3.125000 NA
#> 194 coi[194] 7.292000 7 1.28514778 6 9 4.501414 3.364928 NA
#> 195 coi[195] 6.000000 6 0.81704146 5 7 Inf 3.125000 NA
#> 196 coi[196] 7.077333 7 1.53704356 5 9 4.531202 3.298419 NA
#> 197 coi[197] 6.000000 6 0.00000000 6 6 NA NA NA
#> 198 coi[198] 4.000000 3 1.41515732 3 6 Inf 3.125000 NA
#> 199 coi[199] 1.000000 1 0.00000000 1 1 NA NA NA
#> 200 coi[200] 1.000000 1 0.00000000 1 1 NA NA NA