-
Notifications
You must be signed in to change notification settings - Fork 0
/
5_swed_robustness_test.html
374 lines (349 loc) · 19.5 KB
/
5_swed_robustness_test.html
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
<!DOCTYPE html>
<html xmlns="http://www.w3.org/1999/xhtml">
<head>
<meta charset="utf-8">
<meta http-equiv="Content-Type" content="text/html; charset=utf-8" />
<meta name="generator" content="pandoc" />
<title>Modern Sweden Robustness Test</title>
<script src="library/jquery-1.11.3/jquery.min.js"></script>
<meta name="viewport" content="width=device-width, initial-scale=1" />
<link href="library/bootstrap-3.3.5/css/paper.min.css" rel="stylesheet" />
<script src="library/bootstrap-3.3.5/js/bootstrap.min.js"></script>
<script src="library/bootstrap-3.3.5/shim/html5shiv.min.js"></script>
<script src="library/bootstrap-3.3.5/shim/respond.min.js"></script>
<script src="library/navigation-1.1/tabsets.js"></script>
<style type="text/css">code{white-space: pre;}</style>
<style type="text/css">
div.sourceCode { overflow-x: auto; }
table.sourceCode, tr.sourceCode, td.lineNumbers, td.sourceCode {
margin: 0; padding: 0; vertical-align: baseline; border: none; }
table.sourceCode { width: 100%; line-height: 100%; }
td.lineNumbers { text-align: right; padding-right: 4px; padding-left: 4px; color: #aaaaaa; border-right: 1px solid #aaaaaa; }
td.sourceCode { padding-left: 5px; }
code > span.kw { color: #007020; font-weight: bold; } /* Keyword */
code > span.dt { color: #902000; } /* DataType */
code > span.dv { color: #40a070; } /* DecVal */
code > span.bn { color: #40a070; } /* BaseN */
code > span.fl { color: #40a070; } /* Float */
code > span.ch { color: #4070a0; } /* Char */
code > span.st { color: #4070a0; } /* String */
code > span.co { color: #60a0b0; font-style: italic; } /* Comment */
code > span.ot { color: #007020; } /* Other */
code > span.al { color: #ff0000; font-weight: bold; } /* Alert */
code > span.fu { color: #06287e; } /* Function */
code > span.er { color: #ff0000; font-weight: bold; } /* Error */
code > span.wa { color: #60a0b0; font-weight: bold; font-style: italic; } /* Warning */
code > span.cn { color: #880000; } /* Constant */
code > span.sc { color: #4070a0; } /* SpecialChar */
code > span.vs { color: #4070a0; } /* VerbatimString */
code > span.ss { color: #bb6688; } /* SpecialString */
code > span.im { } /* Import */
code > span.va { color: #19177c; } /* Variable */
code > span.cf { color: #007020; font-weight: bold; } /* ControlFlow */
code > span.op { color: #666666; } /* Operator */
code > span.bu { } /* BuiltIn */
code > span.ex { } /* Extension */
code > span.pp { color: #bc7a00; } /* Preprocessor */
code > span.at { color: #7d9029; } /* Attribute */
code > span.do { color: #ba2121; font-style: italic; } /* Documentation */
code > span.an { color: #60a0b0; font-weight: bold; font-style: italic; } /* Annotation */
code > span.cv { color: #60a0b0; font-weight: bold; font-style: italic; } /* CommentVar */
code > span.in { color: #60a0b0; font-weight: bold; font-style: italic; } /* Information */
</style>
<style type="text/css">
pre:not([class]) {
background-color: white;
}
</style>
<style type = "text/css">
.main-container {
max-width: 940px;
margin-left: auto;
margin-right: auto;
}
code {
color: inherit;
background-color: rgba(0, 0, 0, 0.04);
}
img {
max-width:100%;
height: auto;
}
</style>
<script src="library/auto_tab_first_section.js"></script>
</head>
<body>
<div style="width:100%;height:200px;background-image:url('library/header/swed.jpg');background-size:cover;"></div>
<div class="container-fluid main-container"><div class="row"><div class="col-md-4"><a href="index.html"><i class="glyphicon glyphicon-th-list"></i> Back to index</a></div></div></div>
<div class="container-fluid main-container">
<div id="th-century-sweden-robustness-analyses" class="section level1 tab-content">
<h1>20th century Sweden robustness analyses</h1>
<div id="loading-details" class="section level3 accordion">
<h3>Loading details</h3>
<div class="sourceCode"><pre class="sourceCode r"><code class="sourceCode r"><span class="kw">source</span>(<span class="st">"0__helpers.R"</span>)
opts_chunk$<span class="kw">set</span>(<span class="dt">warning=</span><span class="ot">TRUE</span>, <span class="dt">cache=</span>F,<span class="dt">cache.lazy=</span>F,<span class="dt">tidy=</span><span class="ot">FALSE</span>,<span class="dt">autodep=</span><span class="ot">TRUE</span>,<span class="dt">dev=</span><span class="kw">c</span>(<span class="st">'png'</span>,<span class="st">'pdf'</span>),<span class="dt">fig.width=</span><span class="dv">20</span>,<span class="dt">fig.height=</span><span class="fl">12.5</span>,<span class="dt">out.width=</span><span class="st">'1440px'</span>,<span class="dt">out.height=</span><span class="st">'900px'</span>,<span class="dt">cache.extra=</span><span class="kw">file.info</span>(<span class="st">'swed1.rdata'</span>)[, <span class="st">'mtime'</span>])
make_path =<span class="st"> </span>function(file) {
<span class="kw">get_coefficient_path</span>(file, <span class="st">"swed"</span>)
}
<span class="co"># options for each chunk calling knit_child</span>
opts_chunk$<span class="kw">set</span>(<span class="dt">warning=</span><span class="ot">FALSE</span>, <span class="dt">message =</span> <span class="ot">FALSE</span>, <span class="dt">echo =</span> <span class="ot">FALSE</span>)</code></pre></div>
</div>
<div id="analysis-description" class="section level2">
<h2>Analysis description</h2>
<div id="data-subset" class="section level3">
<h3>Data subset</h3>
<p>The <code>swed_subset_children.1</code> dataset is based on the full dataset of all participants where paternal age is known and birth years are from 1947 to 1959. The subset contains 88707 randomly drawn participants from 56679 families.</p>
</div>
<div id="model-description" class="section level3">
<h3>Model description</h3>
<p>All of the following models are the same as our main model m3, except for the noted changes to test robustness.</p>
</div>
</div>
<div id="r1-relaxed-exclusion-criteria" class="section level2 tab-content">
<h2><em>r1</em>: Relaxed exclusion criteria</h2>
<p>For the four historical populations, we imposed quite stringent exclusion criteria to ensure sufficient data quality for our intended analysis. As this was not necessary for the modern Swedish data, there are no exclusion criteria to relax.</p>
</div>
<div id="r2-fewer-covariates" class="section level2 tab-content">
<h2><em>r2</em>: Fewer covariates</h2>
<p>Adding covariates increases the complexity of the model and makes it harder to interpret. We chose to adjust for many potential confounds because we are interested in causal isolation of the paternal age effect. Here we show what happens when only birth cohort and anchor sex are adjusted for.</p>
<div id="model-summary" class="section level3 tab-content">
<h3>Model summary</h3>
<div id="full-summary" class="section level4">
<h4>Full summary</h4>
<div class="sourceCode"><pre class="sourceCode r"><code class="sourceCode r">model_summary =<span class="st"> </span><span class="kw">summary</span>(model, <span class="dt">use_cache =</span> <span class="ot">FALSE</span>)
<span class="kw">print</span>(model_summary)</code></pre></div>
<pre><code>## Family: poisson (log)
## Formula: children ~ paternalage.diff + birth_cohort + male + paternalage.mean + (1 | idParents)
## Data: model_data (Number of observations: 88707)
## Samples: 6 chains, each with iter = 800; warmup = 300; thin = 1;
## total post-warmup samples = 3000
## WAIC: Not computed
##
## Group-Level Effects:
## ~idParents (Number of levels: 56679)
## Estimate Est.Error l-95% CI u-95% CI Eff.Sample Rhat
## sd(Intercept) 0.01 0.01 0 0.03 247 1.01
##
## Population-Level Effects:
## Estimate Est.Error l-95% CI u-95% CI Eff.Sample Rhat
## Intercept 0.65 0.01 0.64 0.66 3000 1
## paternalage.diff -0.04 0.01 -0.05 -0.02 3000 1
## birth_cohort1950M1955 0.00 0.01 -0.01 0.01 3000 1
## birth_cohort1955M1960 -0.01 0.01 -0.03 0.00 3000 1
## male1 -0.06 0.00 -0.07 -0.05 3000 1
## paternalage.mean -0.03 0.00 -0.04 -0.02 3000 1
##
## Samples were drawn using sampling(NUTS). For each parameter, Eff.Sample
## is a crude measure of effective sample size, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).</code></pre>
</div>
<div id="table-of-fixed-effects" class="section level4">
<h4>Table of fixed effects</h4>
<div class="sourceCode"><pre class="sourceCode r"><code class="sourceCode r">fixed_eff =<span class="st"> </span><span class="kw">data.frame</span>(model_summary$fixed, <span class="dt">check.names =</span> F)
fixed_eff$Est.Error =<span class="st"> </span>fixed_eff$Eff.Sample =<span class="st"> </span>fixed_eff$Rhat =<span class="st"> </span><span class="ot">NULL</span>
fixed_eff$OR =<span class="st"> </span><span class="kw">exp</span>(fixed_eff$Estimate)
fixed_eff$OR_low =<span class="st"> </span><span class="kw">exp</span>(fixed_eff$<span class="st">`</span><span class="dt">l-95% CI</span><span class="st">`</span>)
fixed_eff$OR_high =<span class="st"> </span><span class="kw">exp</span>(fixed_eff$<span class="st">`</span><span class="dt">u-95% CI</span><span class="st">`</span>)
pander::<span class="kw">pander</span>(fixed_eff)</code></pre></div>
<table>
<colgroup>
<col width="32%" />
<col width="12%" />
<col width="12%" />
<col width="12%" />
<col width="8%" />
<col width="10%" />
<col width="10%" />
</colgroup>
<thead>
<tr class="header">
<th align="center"> </th>
<th align="center">Estimate</th>
<th align="center">l-95% CI</th>
<th align="center">u-95% CI</th>
<th align="center">OR</th>
<th align="center">OR_low</th>
<th align="center">OR_high</th>
</tr>
</thead>
<tbody>
<tr class="odd">
<td align="center"><strong>Intercept</strong></td>
<td align="center">0.6519</td>
<td align="center">0.6413</td>
<td align="center">0.6623</td>
<td align="center">1.919</td>
<td align="center">1.899</td>
<td align="center">1.939</td>
</tr>
<tr class="even">
<td align="center"><strong>paternalage.diff</strong></td>
<td align="center">-0.03644</td>
<td align="center">-0.05117</td>
<td align="center">-0.02178</td>
<td align="center">0.9642</td>
<td align="center">0.9501</td>
<td align="center">0.9785</td>
</tr>
<tr class="odd">
<td align="center"><strong>birth_cohort1950M1955</strong></td>
<td align="center">-0.0003772</td>
<td align="center">-0.01225</td>
<td align="center">0.01163</td>
<td align="center">0.9996</td>
<td align="center">0.9878</td>
<td align="center">1.012</td>
</tr>
<tr class="even">
<td align="center"><strong>birth_cohort1955M1960</strong></td>
<td align="center">-0.013</td>
<td align="center">-0.02548</td>
<td align="center">-7.683e-05</td>
<td align="center">0.9871</td>
<td align="center">0.9748</td>
<td align="center">0.9999</td>
</tr>
<tr class="odd">
<td align="center"><strong>male1</strong></td>
<td align="center">-0.05733</td>
<td align="center">-0.06702</td>
<td align="center">-0.04722</td>
<td align="center">0.9443</td>
<td align="center">0.9352</td>
<td align="center">0.9539</td>
</tr>
<tr class="even">
<td align="center"><strong>paternalage.mean</strong></td>
<td align="center">-0.03028</td>
<td align="center">-0.03846</td>
<td align="center">-0.02188</td>
<td align="center">0.9702</td>
<td align="center">0.9623</td>
<td align="center">0.9784</td>
</tr>
</tbody>
</table>
</div>
</div>
<div id="paternal-age-effect" class="section level3">
<h3>Paternal age effect</h3>
<div class="sourceCode"><pre class="sourceCode r"><code class="sourceCode r">pander::<span class="kw">pander</span>(<span class="kw">paternal_age_10y_effect</span>(model))</code></pre></div>
<table>
<colgroup>
<col width="40%" />
<col width="24%" />
<col width="17%" />
<col width="17%" />
</colgroup>
<thead>
<tr class="header">
<th align="center">effect</th>
<th align="center">median_estimate</th>
<th align="center">ci_95</th>
<th align="center">ci_80</th>
</tr>
</thead>
<tbody>
<tr class="odd">
<td align="center">estimate father -5y</td>
<td align="center">1.96</td>
<td align="center">[1.93;1.98]</td>
<td align="center">[1.94;1.97]</td>
</tr>
<tr class="even">
<td align="center">estimate father +5y</td>
<td align="center">1.89</td>
<td align="center">[1.86;1.91]</td>
<td align="center">[1.87;1.9]</td>
</tr>
<tr class="odd">
<td align="center">percentage change</td>
<td align="center">-3.6</td>
<td align="center">[-5;-2.2]</td>
<td align="center">[-4.5;-2.6]</td>
</tr>
<tr class="even">
<td align="center">absolute change for reference individual</td>
<td align="center">-0.07</td>
<td align="center">[-0.1;-0.04]</td>
<td align="center">[-0.09;-0.05]</td>
</tr>
</tbody>
</table>
</div>
<div id="marginal-effect-plots" class="section level3">
<h3>Marginal effect plots</h3>
<div class="sourceCode"><pre class="sourceCode r"><code class="sourceCode r"><span class="kw">plot.brmsMarginalEffects_shades</span>(
<span class="dt">x =</span> <span class="kw">marginal_effects</span>(model, <span class="dt">re_formula =</span> <span class="ot">NA</span>, <span class="dt">probs =</span> <span class="kw">c</span>(<span class="fl">0.025</span>,<span class="fl">0.975</span>)),
<span class="dt">y =</span> <span class="kw">marginal_effects</span>(model, <span class="dt">re_formula =</span> <span class="ot">NA</span>, <span class="dt">probs =</span> <span class="kw">c</span>(<span class="fl">0.1</span>,<span class="fl">0.9</span>)),
<span class="dt">ask =</span> <span class="ot">FALSE</span>)</code></pre></div>
<p><img src="5_swed_robustness_test_files/figure-html/unnamed-chunk-8-1.png" width="1440px" height="900px" /><img src="5_swed_robustness_test_files/figure-html/unnamed-chunk-8-2.png" width="1440px" height="900px" /><img src="5_swed_robustness_test_files/figure-html/unnamed-chunk-8-3.png" width="1440px" height="900px" /><img src="5_swed_robustness_test_files/figure-html/unnamed-chunk-8-4.png" width="1440px" height="900px" /></p>
</div>
<div id="coefficient-plot" class="section level3">
<h3>Coefficient plot</h3>
<p>Coefficient estimates (95% and 80% credibility).</p>
<div class="sourceCode"><pre class="sourceCode r"><code class="sourceCode r"><span class="kw">mcmc_intervals</span>(<span class="kw">as.matrix</span>(model$fit), <span class="dt">regex_pars =</span> <span class="st">"b_[^I]"</span>)</code></pre></div>
<p><img src="5_swed_robustness_test_files/figure-html/unnamed-chunk-9-1.png" width="1440px" height="900px" /></p>
<div class="sourceCode"><pre class="sourceCode r"><code class="sourceCode r"><span class="kw">mcmc_areas</span>(<span class="kw">as.matrix</span>(model$fit), <span class="dt">regex_pars =</span> <span class="st">"b"</span>)</code></pre></div>
<p><img src="5_swed_robustness_test_files/figure-html/unnamed-chunk-9-2.png" width="1440px" height="900px" /></p>
</div>
<div id="diagnostics" class="section level3 tab-content">
<h3>Diagnostics</h3>
<p>These plots were made to diagnose misfit and nonconvergence.</p>
<div id="posterior-predictive-checks" class="section level4">
<h4>Posterior predictive checks</h4>
<div class="sourceCode"><pre class="sourceCode r"><code class="sourceCode r">brms::<span class="kw">pp_check</span>(model, <span class="dt">re_formula =</span> <span class="ot">NA</span>, <span class="dt">type =</span> <span class="st">"dens_overlay"</span>)</code></pre></div>
<p><img src="5_swed_robustness_test_files/figure-html/unnamed-chunk-10-1.png" width="1440px" height="900px" /></p>
<div class="sourceCode"><pre class="sourceCode r"><code class="sourceCode r">brms::<span class="kw">pp_check</span>(model, <span class="dt">re_formula =</span> <span class="ot">NA</span>, <span class="dt">type =</span> <span class="st">"hist"</span>)</code></pre></div>
<p><img src="5_swed_robustness_test_files/figure-html/unnamed-chunk-10-2.png" width="1440px" height="900px" /></p>
</div>
<div id="rhat" class="section level4">
<h4>Rhat</h4>
<p>Did the 6 chains converge?</p>
<div class="sourceCode"><pre class="sourceCode r"><code class="sourceCode r"><span class="kw">stanplot</span>(model, <span class="dt">pars =</span> <span class="st">"^b_[^I]"</span>, <span class="dt">type =</span> <span class="st">'rhat'</span>)</code></pre></div>
<p><img src="5_swed_robustness_test_files/figure-html/unnamed-chunk-11-1.png" width="1440px" height="900px" /></p>
</div>
<div id="effective-sample-size-by-average-sample-size" class="section level4">
<h4>Effective sample size by average sample size</h4>
<div class="sourceCode"><pre class="sourceCode r"><code class="sourceCode r"><span class="kw">stanplot</span>(model, <span class="dt">pars =</span> <span class="st">"^b"</span>, <span class="dt">type =</span> <span class="st">'ess'</span>)</code></pre></div>
<p><img src="5_swed_robustness_test_files/figure-html/unnamed-chunk-12-1.png" width="1440px" height="900px" /></p>
</div>
<div id="monte-carlo-se" class="section level4">
<h4>Monte Carlo SE</h4>
<div class="sourceCode"><pre class="sourceCode r"><code class="sourceCode r"><span class="kw">stanplot</span>(model, <span class="dt">pars =</span> <span class="st">"^b"</span>, <span class="dt">type =</span> <span class="st">'mcse'</span>)</code></pre></div>
<p><img src="5_swed_robustness_test_files/figure-html/unnamed-chunk-13-1.png" width="1440px" height="900px" /></p>
</div>
<div id="trace-plots" class="section level4">
<h4>Trace plots</h4>
<div class="sourceCode"><pre class="sourceCode r"><code class="sourceCode r">if(<span class="kw">any</span>( <span class="kw">summary</span>(model)$fixed[,<span class="st">"Rhat"</span>] ><span class="st"> </span><span class="fl">1.1</span>)) { <span class="co"># only do traceplots if not converged</span>
<span class="kw">plot</span>(model, <span class="dt">N =</span> <span class="dv">3</span>, <span class="dt">ask =</span> <span class="ot">FALSE</span>)
}</code></pre></div>
</div>
<div id="further-plots" class="section level4">
<h4>Further plots</h4>
<div class="sourceCode"><pre class="sourceCode r"><code class="sourceCode r"><span class="kw">stanplot</span>(model, <span class="dt">pars =</span> <span class="st">"^b"</span>, <span class="dt">type =</span> <span class="st">'diag'</span>)</code></pre></div>
<p><img src="5_swed_robustness_test_files/figure-html/unnamed-chunk-15-1.png" width="1440px" height="900px" /></p>
</div>
<div id="file-name" class="section level4">
<h4>File name</h4>
<p>coefs/swed/r2_few_controls.rds</p>
</div>
<div id="cluster-script" class="section level4">
<h4>Cluster script</h4>
<div class="sourceCode"><pre class="sourceCode r"><code class="sourceCode r">clusterscript =<span class="st"> </span><span class="kw">str_replace</span>(<span class="kw">basename</span>(model_filename), <span class="st">"</span><span class="ch">\\</span><span class="st">.rds"</span>,<span class="st">".html"</span>)
<span class="kw">cat</span>(<span class="st">"[Cluster script]("</span> , clusterscript, <span class="st">")"</span>, <span class="dt">sep =</span> <span class="st">""</span>)</code></pre></div>
<p><a href="r2_few_controls.html">Cluster script</a></p>
<div class="sourceCode"><pre class="sourceCode r"><code class="sourceCode r">opts_chunk$<span class="kw">set</span>(<span class="dt">results =</span> <span class="st">"asis"</span>, <span class="dt">echo =</span> <span class="ot">FALSE</span>)</code></pre></div>
</div>
</div>
</div>
</div>
</div>
<script>
// add bootstrap table styles to pandoc tables
$(document).ready(function () {
$('tr.header').parent('thead').parent('table').addClass('table table-condensed');
});
</script>
</body>
</html>