Compare different regtregression nahods</e>Comp<titlscript>// Pc" /> 2.9 adds attributes on both > er c" ferv. Wetremove the former (to // betenible" co with the behavior of Pc" /> < 2.8). />umreg.addEvregListator('DOMCnt="20Lo ed', function(e) { var hs = />umreg.querySelec coAll("erv.section[class*='level'] > :first-child"); var i, nita; for (i = 0; i < hs.length; i++) { h = hs[i]; if (!/^h[1-6]$/i.test(h.tagN"da))tent=inue; // it should beta > er h1-h6 a = h.attributes; while (a.length > 0) h.removeAttribute(a[0].="da); } }); </script><titlstyle typatetext/css"><tcode{white-space: prdthrap;} span.smallcaps{fnt=-variant: small-caps;} span.un erline{text-decor" ion: un erline;} erv.column{display:tialine-block; verti=1"-align: top; h, in: 50%;} erv.hanging-in ent{margin-left: 1.5em; text-in ent: -1.5em;} ul.task-list{list-style: none;} </style><tit<titlstyle typatetext/css"><tcode { white-space: prd; } .sourceCode { overflow: vis" co; } </style><tlstyle typatetext/css" " ca-originndoc" /> ><tprd > code.sourceCode { white-space: prd; posi ion: rel" ivd; }<tprd > code.sourceCode > span { line-height: 1.25; }<tprd > code.sourceCode > span:empty { height: 1.2em; } .sourceCode { overflow: vis" co; } code.sourceCode > span { color:tiaherit; text-decor" ion: iaherit; } erv.sourceCode { margin: 1em 0; }<tprd.sourceCode { margin: 0; }<t@media screen { erv.sourceCode { overflow: or"o; }<t}<t@media priegt{<tprd > code.sourceCode { white-space: prdthrap; }<tprd > code.sourceCode > span { display:tialine-block; text-in ent: -5em; padding-left: 5em; } }<tprd.numberSource code { cout="r-reset: source-line 0; }<tprd.numberSource code > span { posi ion: rel" ivd; left: -4em; cout="r-increment: source-line; }<tprd.numberSource code > span > a:first-child::before { cot="20: cout="r(source-line); posi ion: rel" ivd; left: -1em; text-align: right; verti=1"-align: baseline; bor er: none; display:tialine-block; -webkit-touch-callout: none; -webkit-us"r-selec : none; -k> -us"r-selec : none; -moz-us"r-selec : none; -ms-us"r-selec : none; us"r-selec : none; padding: 0 4px; h, in: 4em; color:t#aaaaaa; }<tprd.numberSource { margin-left: 3em; bor er-left: 1px solidt#aaaaaa; padding-left: 4px; } erv.sourceCode { }<t@media screen { prd > code.sourceCode > span > a:first-child::before { text-decor" ion: un erline; } }<tcode span.al { color:t#ff0000; fnt=-weight: bold; } <tcode span.an { color:t#60a0b0; fnt=-weight: bold; fnt=-style: italic; } <tcode span.at { color:t#7d9029; } <tcode span.bn { color:t#40a070; } <tcode span.bu { color:t#008000; } <tcode span.cf { color:t#007020; fnt=-weight: bold; } <tcode span.ch { color:t#4070a0; } <tcode span.cn { color:t#880000; } <tcode span.co { color:t#60a0b0; fnt=-style: italic; } <tcode span.cv { color:t#60a0b0; fnt=-weight: bold; fnt=-style: italic; } <tcode span.do { color:t#ba2121; fnt=-style: italic; } <tcode span.dt { color:t#902000; } <tcode span.dv { color:t#40a070; } <tcode span.er { color:t#ff0000; fnt=-weight: bold; } <tcode span.ex { } <tcode span.fl { color:t#40a070; } <tcode span.fu { color:t#06287e; } <tcode span.im { color:t#008000; fnt=-weight: bold; } <tcode span.in { color:t#60a0b0; fnt=-weight: bold; fnt=-style: italic; } <tcode span.kw { color:t#007020; fnt=-weight: bold; } <tcode span.op { color:t#666666; } <tcode span.ot { color:t#007020; } <tcode span.pp { color:t#bc7a00; } <tcode span.sc { color:t#4070a0; } <tcode span.ss { color:t#bb6688; } <tcode span.st { color:t#4070a0; } <tcode span.va { color:t#19177c; } <tcode span.vs { color:t#4070a0; } <tcode span.wa { color:t#60a0b0; fnt=-weight: bold; fnt=-style: italic; } <t</style><tlscript><t// apply oc" /> erv.sourceCode style to prd.sourceCode instaad (function() { var sheets = />umreg.styleSheets; for (var i = 0; i < sheets.length; i++) { if (sheets[i].owtorNode." caset["origin"] !== doc" /> )tent=inue; try { var rules = sheets[i].cssRules; } catch (e) {tent=inue; } var j = 0; while (j < rules.length) { var rule = rules[j]; // check if there is a erv.sourceCode rule if (rule.typa !== rule.STYLE_RULE || rule.selec coText !== derv.sourceCode") { j++; ent=inue; } var style = rule.style.cssText; // check if color or backgroutd-color is set if (rule.style.color === '' && rule.style.backgroutdColor === '') { j++; ent=inue; } // replace erv.sourceCode by a prd.sourceCode rule sheets[i].deleteRule(j); sheets[i].insertRule('prd.sourceCode{' + style + '}', j); } } })(); </script><titit<titlstyle typatetext/css">body { backgroutd-color:t#fff; margin: 1em or"o; max-h, in: 700px; overflow: vis" co; padding-left: 2em; padding-right: 2em; fnt=-family:t"Open Sans", "Helveti=1 Neue", Helveti=1, Arial, sans-serif; fnt=-size: 14px; line-height: 1.35; }<t#TOC { clear: both; margin: 0 0 10px 10px; padding: 4px; h, in: 400px; bor er: 1px solidt#CCCCCC; bor er-radius: 5px; backgroutd-color:t#f6f6f6; fnt=-size: 13px; line-height: 1.3; }<t#TOC .toce>Com { fnt=-weight: bold; fnt=-size: 15px; margin-left: 5px; }<t#TOC ul { padding-left: 40px; margin-left: -1.5em; margin-top: 5px; margin-bottom: 5px; }<t#TOC ul ul { margin-left: -2em; }<t#TOC li { line-height: 16px; }<ttabom { margin: 1em or"o; bor er-h, in: 1px; bor er-color:t#DDDDDD; bor er-style: outset; bor er-collapse: collapse; }<ttabom th { bor er-h, in: 2px; padding: 5px; bor er-style: inset; }<ttabom td { bor er-h, in: 1px; bor er-style: inset; line-height: 18px; padding: 5px 5px; }<ttabom, tabom th, tabom td { bor er-left-style: none; bor er-right-style: none; }<ttabom thaad, tabom tr.even { backgroutd-color:t#f7f7f7; }<tp { margin: 0.5em 0; }<tblockquote { backgroutd-color:t#f6f6f6; padding: 0.25em 0.75em; }<thr { bor er-style: solid; bor er: none; bor er-top: 1px solidt#777; margin: 28px 0; }<tdl { margin-left: 0; }<tdl dd { margin-bottom: 13px; margin-left: 13px; }<tdl dt { fnt=-weight: bold; } ul { margin-top: 0; } ul li { list-style: circle outside; } ul ul { margin-bottom: 0; }<tprd, code { backgroutd-color:t#f7f7f7; bor er-radius: 3px; color:t#333; white-space: prdthrap; }<tprd { bor er-radius: 3px; margin: 5px 0px 10px 0px; padding: 10px; }<tprd:not([class]) { backgroutd-color:t#f7f7f7; }<tcode { fnt=-family:tConsolas, Monaco, 'Courier New', monospace; fnt=-size: 85%; }<tp > code, li > code { padding: 2px 0px; } erv.figurd { text-align: cet="r; } img { backgroutd-color:t#FFFFFF; padding: 2px; bor er: 1px solidt#DDDDDD; bor er-radius: 3px; bor er: 1px solidt#CCCCCC; margin: 0 5px; }<th1 { margin-top: 0; fnt=-size: 35px; line-height: 40px; } h2 { bor er-bottom: 4px solidt#f7f7f7; padding-top: 10px; padding-bottom: 2px; fnt=-size: 145%; }<th3 { bor er-bottom: 2px solidt#f7f7f7; padding-top: 10px; fnt=-size: 120%; }<th4 { bor er-bottom: 1px solidt#f7f7f7; margin-left: 8px; fnt=-size: 105%; }<th5, h6 { bor er-bottom: 1px solidt#ccc; fnt=-size: 105%; }<ta { color:t#0033dd; text-decor" ion: none; }<ta:hover { color:t#6666ff; }<ta:vis"ted { color:t#800080; }<ta:vis"ted:hover { color:t#BB00BB; }<ta[href^="-equ:"] { text-decor" ion: un erline; } a[href^="-equs:"] { text-decor" ion: un erline; } <tcode > span.kw { color:t#555; fnt=-weight: bold; } <tcode > span.dt { color:t#902000; } <tcode > span.dv { color:t#40a070; } <tcode > span.bn { color:t#d14; } <tcode > span.fl { color:t#d14; } <tcode > span.ch { color:t#d14; } <tcode > span.st { color:t#d14; } <tcode > span.co { color:t#888888; fnt=-style: italic; } <tcode > span.ot { color:t#007020; } <tcode > span.al { color:t#ff0000; fnt=-weight: bold; } <tcode > span.fu { color:t#900; fnt=-weight: bold; } <tcode > span.er { color:t#a61717; backgroutd-color:t#e3d2d2; } <t</style><titit<titl/> <metabody><titit<titlh1 classtet>Com toc-ignore"pare different regtregression nahods</h1><tlh4 classteor" con>ej Jevsenak" /></h4><tlh4 classte" con>-07-18" /></h4><tit<titlstyle> body { text-align: justify; fnt=-size: 16px} </style><tlerv idtethe-cre dif_ nahods" classtesection level2"><tlh2>1. The <em>cre dif_ nahods()</em></h2><tThis is a s cot vignette about the <em>cre dif_ nahods()</em> function from the dendroTools R package. The <em>cre dif_ nahods()</em> us"s k-fold cross-valid" ion to cre different regtregression nahods. In addi ion, nahods could betcre difd by 1) usingta >oldout " ca, where e ca is excludfd from the hypar dia naer optimiz" ion and<tcross-valid" ion c" f nahods difflaaer addi ionally evaluated for ahose points, or 2) similarly by usingtedge " ca, where extreme (edge) points are estimated as a valid" ion " ca. Cur regly, there are fivetregression mnahods implemented: artificial neural networks with the Bayesian regulariz" ion trainingtalgorithm (BRNN), (ensembom of) model tre"s (MT), rc" /m forests of regression tre"s (RF) c" f(multiple) linear regression (MLR). The calculated parformancef narics diffthe correl" ion coefficiregt(r), the root mean squdifd error (RMSE), the root rel" ivd squdifd error (RRSE), the in ex of agreement (d), the reduction of error (RE), the coefficiregtof efficirecy (CE), the dnarendfd efficirecy (DE) a" f nan bias, calculated as the drent rece between observed and<testimated nan response for ahe valid" ion c" fcalibr" ion " ca. The outpugtof the <em>cre dif_ nahods()</em> function is a list with 18<telements, which could betrnarievfd by callingtthe “$” opor" co c" fthe elemente="da. Please note, the examples prdsented here are made cre uta ionally less it="2sivetto sa isfy the CRAN policy.itleabom><tlt> <m<tr><tlt> styletetext-align:left;"><tElement </th><tlt> styletetext-align:left;"><tElement_description </th><tl/tr><tl/t> <m<tbody><t<tr><tltd styletetext-align:left;"><t$ nan_std<tl/td><tltd styletetext-align:left;"><te ca fia n with calculated narics for ahe selec ed regression nahods. For each regression nahod c" feach calculated naric, nan c" fstc" ard<tce-w" ion cre given<tl/td><tl/tr><tltr><tltd styletetext-align:left;"><t$std_rc"ks<tl/td><tltd styletetext-align:left;"><te ca fia n with rc"kstof calculated narics: nan rc"k c" fshcre of rc"k_1 cre given<tl/td><tl/tr><tltr><tltd styletetext-align:left;"><t$edge_results<tl/td><tltd styletetext-align:left;"><te ca fia n with calculated parformancef narics for ahe cet=ral-edge test. The cet=ral pargtof the e ca reprdsents the calibr" ion " ca, whiom thatedge " ca, i.e. extreme values, reprdsent ahe valid" ion " ca. Drent regtregression odels diffcalibr" ed usingtahe cet=ral e ca and<tvalid" ed for ahatedge (extreme) " ca. This test is pargicularly im" coant ao assess ahatparformancefof odels for ahatprddiction of the extreme " ca. Thefshcre of ahatedge (extreme) " ca is defined with the edge_shcre argument </td><tl/tr><tltr><tltd styletetext-align:left;"><t$>oldout_results<tl/td><tltd styletetext-align:left;"><tcalculated narics for ahe >oldout " ca </td><tl/tr><tltr><tltd styletetext-align:left;"><t$bias_cal<tl/td><tltd styletetext-align:left;"><tggplot objec fof nan bias for calibr" ion " ca </td><tl/tr><tltr><tltd styletetext-align:left;"><t$bias_val<tl/td><tltd styletetext-align:left;"><tggplot objec fof nan bias for valid" ion " ca </td><tl/tr><tltr><tltd styletetext-align:left;"><t$=ransnt _functions<tl/td><tltd styletetext-align:left;"><tggplot or plotly objec fwith transnt functionstof erent regt nahods, facet is us"dtto se diatet nahods </td><tl/tr><tltr><tltd styletetext-align:left;"><t$=ransnt _functions_together<tl/td><tltd styletetext-align:left;"><tggplot or plotly objec fwith transnt functionstof mnahods plotted together<tl/td><tl/tr><tltr><tltd styletetext-align:left;"><t$ dia naer_values<tl/td><tltd styletetext-align:left;"><ta e ca fia n with specificationstof dia naers us"dtfor erent reg regression nahods </td><tl/tr><tltr><tltd styletetext-align:left;"><t$PCA_outpug<tl/td><tltd styletetext-align:left;"><tpriecre objec : the result outpugtof the PCA analysis </td><tl/tr><tltr><tltd styletetext-align:left;"><t$reconstructions<tl/td><tltd styletetext-align:left;"><tggplot objec : reconstruct"dtdepen ent variabom based on the " caset_cre lete argument, facet is us"dtto split plots by nahods </td><tl/tr><tltr><tltd styletetext-align:left;"><t$reconstructions_together<tl/td><tltd styletetext-align:left;"><tggplot objec : reconstruct"dtdepen ent variabom based on the " caset_cre lete argument, all reconstructions diffon the sa n plot </td><tl/tr><tltr><tltd styletetext-align:left;"><t$normal_QQ_cal<tl/td><tltd styletetext-align:left;"><tnormal q-q plot for calibr" ion " ca </td><tl/tr><tltr><tltd styletetext-align:left;"><t$normal_QQ_>oldout<tl/td><tltd styletetext-align:left;"><tnormal q-q plot for >oldout " ca </td><tl/tr><tltr><tltd styletetext-align:left;"><t$normal_QQ_edge l/td><tltd styletetext-align:left;"><tnormal q-q plot for edge " ca </td><tl/tr><tltr><tltd styletetext-align:left;"><t$residuals_vs_fitted_cal<tl/td><tltd styletetext-align:left;"><tresiduals vs fitted values plot for calibr" ion " ca </td><tl/tr><tltr><tltd styletetext-align:left;"><t$residuals_vs_fitted_>oldout<tl/td><tltd styletetext-align:left;"><tresiduals vs fitted values plot for >oldout " ca </td><tl/tr><tltr><tltd styletetext-align:left;"><t$residuals_vs_fitted_edge l/td><tltd styletetext-align:left;"><tresiduals vs fitted values plot for edge " ca </td><tl/tr><tl/tbody><t</eabom><tl/div><tlerv idtebasic-example" classtesection level2"><tlh2>2. Basic example</h2><t<p>For ahe basic example, wn will us" the e caset with the Mnan Vessel<tArea (MVA) chronology c" fthe nan April tempor" ure, the e caset is saved as " caset_MVA. All fivetregression mnahods will betcre difd with 10-fold cross-valid" ion repeated 1 times. The <em>optimize</em> argument is settto TRUE, therefore all tuningt dia naers will betdefined in a prdliminary optimiz" ion phase. Afaer the coe diison, the outpug<telements diffrnarievfd with the “$” opor" co.</p><tlerv classtesourceCode" idtecb1"><prd classtesourceCode r"><code classtesourceCode r"><span idtecb1-1"><a href="#cb1-1" tabin ex="-1"></a><span classteco"># Lo the dendroTools R package</span></span>itlspan idtecb1-2"><a href="#cb1-2" tabin ex="-1"></a><span classtefu">libr"ry</span>(dendroTools)</span>itlspan idtecb1-3"><a href="#cb1-3" tabin ex="-1"></a></span>itlspan idtecb1-4"><a href="#cb1-4" tabin ex="-1"></a><span classteco"># Lo the d ca</span></span>itlspan idtecb1-5"><a href="#cb1-5" tabin ex="-1"></a><span classtefu">d ca</span>(" caset_MVA)</span>itlspan idtecb1-6"><a href="#cb1-6" tabin ex="-1"></a></span>itlspan idtecb1-7"><a href="#cb1-7" tabin ex="-1"></a><span classteco"># Basic example</span></span>itlspan idtecb1-8"><a href="#cb1-8" tabin ex="-1"></a>basic_example <span classteot"><-</span> <span classtefu">cre dif_ nahods</span>(<span classteat">formula =</span> T_Apr <span classtesc">~</span> MVA, <span classteat">e caset =</span> " caset_MVA, <span classteat">k =</span> <span classtedv">10</span>, <span classteat">repeats =</span> <span classtedv">1</span>, <span classteat">optimize =</span> <span classtecn">TRUE</span>, <span classteat">MT_cremittees_vec co =</span> <span classtefu">c</span>(<span classtedv">1</span>), <span classteat">RF_maxnodes_vec co =</span> <span classtefu">c</span>(<span classtedv">5</span>), <span classteat">RF_nodesize_vec co =</span> <span classtefu">c</span>(<span classtedv">10</span>))</span></code></prd>l/div><tlerv classtesourceCode" idtecb2"><prd classtesourceCode r"><code classtesourceCode r"><span idtecb2-1"><a href="#cb2-1" tabin ex="-1"></a><span classteco"># The e ca fia n with nan c" fstc" ard ce-w" ion of arformancef narics for ahe calibr" ion c" fthe valid" ion " ca</span></span>itlspan idtecb2-2"><a href="#cb2-2" tabin ex="-1"></a><span classtefu">kabom</span>(basic_example<span classtesc">$</span> nan_std)</span></code></prd>l/div><tleabom><tlcolgroup><tlcol h, in="3% <metacol h, in="8% <metacol h, in="8% <metacol h, in="12% <metacol h, in="10% <metacol h, in="10% <metacol h, in="9% <metacol h, in="9% <metacol h, in="8% <metacol h, in="9% <metacol h, in="8% <meta/colgroup><tlt> <m<tr classte> er"><tlt> align="left"></th><tlt> align="left">Mnaric</th><tlt> align="left">Period</th><tlt> align="right">Mnan BRNN</th><tlt> align="right">Std BRNN</th><tlt> align="right">Mnan MLR</th><tlt> align="right">Std MLR</th><tlt> align="right">Mnan MT</th><tlt> align="right">Std MT</th><tlt> align="right">Mnan RF</th><tlt> align="right">Std RF</th><tl/tr><tl/t> <m<tbody><t<tr classteodd"><tltd align="left">1l/td><tltd align="left">rl/td><tltd align="left">call/td><tltd align="right">0.607l/td><tltd align="right">0.026l/td><tltd align="right">0.596l/td><tltd align="right">0.026l/td><tltd align="right">0.596l/td><tltd align="right">0.026l/td><tltd align="right">0.712l/td><tltd align="right">0.016</td><tl/tr><tltr classteeven"><tltd align="left">2l/td><tltd align="left">rl/td><tltd align="left">vall/td><tltd align="right">0.632l/td><tltd align="right">0.286l/td><tltd align="right">0.635l/td><tltd align="right">0.278l/td><tltd align="right">0.635l/td><tltd align="right">0.278l/td><tltd align="right">0.567l/td><tltd align="right">0.321</td><tl/tr><tltr classteodd"><tltd align="left">3l/td><tltd align="left">RMSEl/td><tltd align="left">call/td><tltd align="right">1.162l/td><tltd align="right">0.036l/td><tltd align="right">1.174l/td><tltd align="right">0.036l/td><tltd align="right">1.174l/td><tltd align="right">0.036l/td><tltd align="right">1.032l/td><tltd align="right">0.028</td><tl/tr><tltr classteeven"><tltd align="left">4l/td><tltd align="left">RMSEl/td><tltd align="left">vall/td><tltd align="right">1.186l/td><tltd align="right">0.326l/td><tltd align="right">1.193l/td><tltd align="right">0.335l/td><tltd align="right">1.193l/td><tltd align="right">0.335l/td><tltd align="right">1.235l/td><tltd align="right">0.318</td><tl/tr><tltr classteodd"><tltd align="left">5l/td><tltd align="left">RRSEl/td><tltd align="left">call/td><tltd align="right">0.794l/td><tltd align="right">0.020l/td><tltd align="right">0.802l/td><tltd align="right">0.020l/td><tltd align="right">0.802l/td><tltd align="right">0.020l/td><tltd align="right">0.706l/td><tltd align="right">0.017</td><tl/tr><tltr classteeven"><tltd align="left">6l/td><tltd align="left">RRSEl/td><tltd align="left">vall/td><tltd align="right">0.879l/td><tltd align="right">0.219l/td><tltd align="right">0.884l/td><tltd align="right">0.221l/td><tltd align="right">0.884l/td><tltd align="right">0.221l/td><tltd align="right">0.921l/td><tltd align="right">0.244</td><tl/tr><tltr classteodd"><tltd align="left">7l/td><tltd align="left">dl/td><tltd align="left">call/td><tltd align="right">0.716l/td><tltd align="right">0.024l/td><tltd align="right">0.709l/td><tltd align="right">0.025l/td><tltd align="right">0.709l/td><tltd align="right">0.025l/td><tltd align="right">0.791l/td><tltd align="right">0.017</td><tl/tr><tltr classteeven"><tltd align="left">8l/td><tltd align="left">dl/td><tltd align="left">vall/td><tltd align="right">0.669l/td><tltd align="right">0.149l/td><tltd align="right">0.664l/td><tltd align="right">0.147l/td><tltd align="right">0.664l/td><tltd align="right">0.147l/td><tltd align="right">0.653l/td><tltd align="right">0.165</td><tl/tr><tltr classteodd"><tltd align="left">10l/td><tltd align="left">REl/td><tltd align="left">vall/td><tltd align="right">0.237l/td><tltd align="right">0.426l/td><tltd align="right">0.229l/td><tltd align="right">0.432l/td><tltd align="right">0.230l/td><tltd align="right">0.431l/td><tltd align="right">0.156l/td><tltd align="right">0.530l/td><tl/tr><tltr classteeven"><tltd align="left">12l/td><tltd align="left">CEl/td><tltd align="left">vall/td><tltd align="right">0.183l/td><tltd align="right">0.420l/td><tltd align="right">0.175l/td><tltd align="right">0.427l/td><tltd align="right">0.175l/td><tltd align="right">0.427l/td><tltd align="right">0.097l/td><tltd align="right">0.524</td><tl/tr><tltr classteodd"><tltd align="left">14l/td><tltd align="left">DEl/td><tltd align="left">vall/td><tltd align="right">-0.300l/td><tltd align="right">1.053l/td><tltd align="right">-0.305l/td><tltd align="right">1.027l/td><tltd align="right">-0.305l/td><tltd align="right">1.027l/td><tltd align="right">-0.500l/td><tltd align="right">1.600</td><tl/tr><tl/tbody><t</eabom><tlerv classtesourceCode" idtecb3"><prd classtesourceCode r"><code classtesourceCode r"><span idtecb3-1"><a href="#cb3-1" tabin ex="-1"></a><span classteco"># The e ca fia n with non- dia naric estimation of erent regt nahods: average rc"k c" fshcre of rc"k one</span></span>itlspan idtecb3-2"><a href="#cb3-2" tabin ex="-1"></a><span classtefu">kabom</span>(basic_example<span classtesc">$</span>rc"k)</span></code></prd>l/div><tleabom styleteh, in:100%;"><tlcolgroup><tlcol h, in="2% <metacol h, in="6% <metacol h, in="6% <metacol h, in="12% <metacol h, in="11% <metacol h, in="11% <metacol h, in="10% <metacol h, in="10% <metacol h, in="9% <metacol h, in="10% <metacol h, in="9% <meta/colgroup><tlt> <m<tr classte> er"><tlt> align="left"></th><tlt> align="left">Mnaric</th><tlt> align="left">Period</th><tlt> align="right">Avg rc"k BRNN</th><tlt> align="right">%rc"k_1 BRNN</th><tlt> align="right">Avg rc"k MLR</th><tlt> align="right">%rc"k_1 MLR</th><tlt> align="right">Avg rc"k MT</th><tlt> align="right">%rc"k_1 MT</th><tlt> align="right">Avg rc"k RF</th><tlt> align="right">%rc"k_1 RF</th><tl/tr><tl/t> <m<tbody><t<tr classteodd"><tltd align="left">13l/td><tltd align="left">rl/td><tltd align="left">call/td><tltd align="right">2.0l/td><tltd align="right">0.0l/td><tltd align="right">3.3l/td><tltd align="right">0.0l/td><tltd align="right">3.6l/td><tltd align="right">0.0l/td><tltd align="right">1.0l/td><tltd align="right">1.0l/td><tl/tr><tltr classteeven"><tltd align="left">14l/td><tltd align="left">rl/td><tltd align="left">vall/td><tltd align="right">2.4l/td><tltd align="right">0.2l/td><tltd align="right">1.9l/td><tltd align="right">0.3l/td><tltd align="right">2.3l/td><tltd align="right">0.3l/td><tltd align="right">3.4l/td><tltd align="right">0.2l/td><tl/tr><tltr classteodd"><tltd align="left">7l/td><tltd align="left">RMSEl/td><tltd align="left">call/td><tltd align="right">2.0l/td><tltd align="right">0.0l/td><tltd align="right">3.0l/td><tltd align="right">0.0l/td><tltd align="right">4.0l/td><tltd align="right">0.0l/td><tltd align="right">1.0l/td><tltd align="right">1.0l/td><tl/tr><tltr classteeven"><tltd align="left">8l/td><tltd align="left">RMSEl/td><tltd align="left">vall/td><tltd align="right">2.3l/td><tltd align="right">0.3l/td><tltd align="right">2.1l/td><tltd align="right">0.4l/td><tltd align="right">2.3l/td><tltd align="right">0.1l/td><tltd align="right">3.3l/td><tltd align="right">0.2l/td><tl/tr><tltr classteodd"><tltd align="left">9l/td><tltd align="left">RRSEl/td><tltd align="left">call/td><tltd align="right">2.0l/td><tltd align="right">0.0l/td><tltd align="right">3.0l/td><tltd align="right">0.0l/td><tltd align="right">4.0l/td><tltd align="right">0.0l/td><tltd align="right">1.0l/td><tltd align="right">1.0l/td><tl/tr><tltr classteeven"><tltd align="left">10l/td><tltd align="left">RRSEl/td><tltd align="left">vall/td><tltd align="right">2.3l/td><tltd align="right">0.3l/td><tltd align="right">2.1l/td><tltd align="right">0.4l/td><tltd align="right">2.3l/td><tltd align="right">0.1l/td><tltd align="right">3.3l/td><tltd align="right">0.2l/td><tl/tr><tltr classteodd"><tltd align="left">11l/td><tltd align="left">dl/td><tltd align="left">call/td><tltd align="right">2.2l/td><tltd align="right">0.0l/td><tltd align="right">3.4l/td><tltd align="right">0.0l/td><tltd align="right">3.4l/td><tltd align="right">0.0l/td><tltd align="right">1.0l/td><tltd align="right">1.0l/td><tl/tr><tltr classteeven"><tltd align="left">12l/td><tltd align="left">dl/td><tltd align="left">vall/td><tltd align="right">2.1l/td><tltd align="right">0.3l/td><tltd align="right">2.6l/td><tltd align="right">0.1l/td><tltd align="right">2.4l/td><tltd align="right">0.3l/td><tltd align="right">2.9l/td><tltd align="right">0.3l/td><tl/tr><tltr classteodd"><tltd align="left">6l/td><tltd align="left">REl/td><tltd align="left">vall/td><tltd align="right">2.3l/td><tltd align="right">0.3l/td><tltd align="right">2.1l/td><tltd align="right">0.4l/td><tltd align="right">2.3l/td><tltd align="right">0.1l/td><tltd align="right">3.3l/td><tltd align="right">0.2l/td><tl/tr><tltr classteeven"><tltd align="left">2l/td><tltd align="left">CEl/td><tltd align="left">vall/td><tltd align="right">2.3l/td><tltd align="right">0.3l/td><tltd align="right">2.1l/td><tltd align="right">0.4l/td><tltd align="right">2.3l/td><tltd align="right">0.1l/td><tltd align="right">3.3l/td><tltd align="right">0.2l/td><tl/tr><tltr classteodd"><tltd align="left">4l/td><tltd align="left">DEl/td><tltd align="left">vall/td><tltd align="right">2.3l/td><tltd align="right">0.3l/td><tltd align="right">2.1l/td><tltd align="right">0.4l/td><tltd align="right">2.3l/td><tltd align="right">0.1l/td><tltd align="right">3.3l/td><tltd align="right">0.2l/td><tl/tr><tl/tbody><t</eabom><tlerv classtesourceCode" idtecb4"><prd classtesourceCode r"><code classtesourceCode r"><span idtecb4-1"><a href="#cb4-1" tabin ex="-1"></a><span classteco"># Se" the histogramfof nan bias for the valid" ion " ca</span></span>itlspan idtecb4-2"><a href="#cb4-2" tabin ex="-1"></a>basic_example<span classtesc">$</span>bias_val</span></code></prd>l/div><tlerv classtefigurd" styletetext-align: cet="r"><tlimg roleteimg" aria-label="Histogramffor the valid" ion " caffor the basic_example" srcte" ca:image/png;base64,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" alt="Histogramffor the valid" ion " caffor the basic_example" <metap classtecaption"><tHistogramffor the valid" ion " caffor the basic_example<t</p><tl/div><tlerv classtesourceCode" idtecb5"><prd classtesourceCode r"><code classtesourceCode r"><span idtecb5-1"><a href="#cb5-1" tabin ex="-1"></a><span classteco"># Se" the histogramfof nan bias for the calibr" ion " ca</span></span>itlspan idtecb5-2"><a href="#cb5-2" tabin ex="-1"></a>basic_example<span classtesc">$</span>bias_cal</span></code></prd>l/div><tlerv classtefigurd" styletetext-align: cet="r"><tlimg roleteimg" aria-label="Histogramffor the calibr" ion " caffor the basic_example" srcte" ca:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAqAAAAHgCAMAAABNUi8GAAABOFBMVEUAAAAAADoAAGYAOmYAOpAAZrYZGT8ZGWIZP4EZYp8aGhozMzM6AAA6OgA6Ojo6OmY6ZmY6ZpA6ZrY6kLY6kNs/GRk/Pz8/gb1NTU1NTW5NTY5NbqtNjshZWVliGRlin9lmAABmOgBmOjpmZgBmZjpmZmZmZpBmkLZmtv9uTU1uTW5uTY5ubm5uq+SBPxmBgb2BvdmOTU2OyP+QOgCQZjqQZmaQkJCQtraQttuQ29uQ2/+fYhmf2dmrbk2r5P+2ZgC2Zjq2kDq2kJC2tpC2tra227a229u22/+2/9u2//+9gT+92dnIjk3I///Zn2LZvYHZ2Z/Z2b3Z2dnbkDrbtpDbtrbb27bb29vb2//b///kq27k///r6+v/tmb/yI7/25D/27b/29v/5Kv//7b//8j//9v//+T///8G015bAAAACXBIWXMAAA7DAAAOwwHHb6hkAAAZP0lEQVR4nO2d/3/bRJrH3ZZvdXNwvbsSF7q0cMtCF1JIWe72NqRh93r0CKGlZQ+2hrolTeL//z84WZKlkSUrmtHzeMbj9/uHvpw2/uix9e5IGs1oBlOAgBn4LgCgDQSFoEFQCBoEhaBBUAgaBIWgcRR0PBpdv1f8dAVAml6Cjt/5fjopDb3S9rt2/BpgkuRhJsjP5y3qypOl9BL0ZHdnOj3bv1FsxyWkmSB3IIIqRWkJmoKgLgT5+eIUdJwd4tPzhV/d+d359EiXYeC7gFhRFHQyGu2U/xGcLZ9OOwjqFkwLGn6U8iH+3V/m23EOQdAAkiIV1LiMR9DOIKiJrqDHt+/Ot+MegqD+k+ITNFNz8s738+24hOQgqPek+ASdHiWnn2f7xVUSgnYGQU30DvFHo9HobvETgnYGQU10z0GN7fR4L4J6T0LQNhDUexKCtoGg3pMQtA0E9Z6EoG0gqPckBG0DQb0nIWgbCOo9aQME7THiqoOgYqO7XGG4nRK0oDLQgipFIagMCKoUhaAyIKhSFILKgKBKUQgqA4IqRSGoDAiqFIWgMiCoUhSCyoCgSlGpoD+8Nyx4/SsEdQBBlaJyQW8mfx7O3HyIoE4gqFKUmqBn+yMe3OAEgppoCXq2f/3edDy6UWzHJSQHQb0nxSdoNu14zLRjexDURPcclCeLOICgJqmgj+9szwR97fPkjze+ExT0KG1BLZ9u18HHGnrDvTrCcDslsn7Qh4mbj+8MLydt6bZgP+hk5PLgBhdB3eqjBQ0/Ku+ofzgcDi/PukNLP/sLOimvkRC0Owhqovp80LKXCUG7g6AmeoKOTT8RtDsIaqJ2q3NsPJhpiqAWIKiJ0YLuXX4i2Q+6U/kZQTuDoCZago5HKS79oAgaXlJ8gta20/1XETS8pBAEfXwHQd1BUKWoUtDDIYK6g6BKUYWge8O3Z3c5DxHUCQRVisoFfXRtuJ02nnui9+KN7XT/VQQNL8m3oIfD4c30JPTRtZtSHfUL2+n+qwgaXpL3jvq03ZS+F1/ZTvdfRdDwknwL+tmTBnw93c5FUL3hXh1huJ0SAU47pgUNL8l3CxrUtGMEDS/J90VSwd5QcsCyuZ3uv4qg4SUFImjSkpYX8QhqA4IqRZmCPjSO7whqB4IqRRmCHg7LTnoEtQRBlaIKQR/fMftAEdQSBFWKmgtaPf1EUFsQVCmqmNVZOf1EUFsQVClqfi/+8qKeCGoFgipFaXbUH3/yffEaQTuDoCaKtzpPdt9BUAcQ1ERP0MlohKAuIKiJ2iF+MtqZIKgLCGqieS8eQZ1AUBPNe/FzQVfw+MUa56cKjln8lfGgamjei19hC1rj/NQpLeg6RGnei0dQJxDURPNePII6gaAmmvfiEdQJBDXRvBePoE4gqInmvXgEdQJBTSKZNIegukm+W9BGENQCBFWKQlAZEFQpCkFlQFClKASVAUGVohBUBgRVikJQGRBUKQpBZUBQpaiVCdp9gJWIoOenCg4J+5XhdmrQgspAC6oUhaAyIKhSFILKgKBKUQgqA4IqRSGoDAiqFIWgMiCoUhSCyoCgSlEIKgOCKkUhqAwIqhSl+mym6/fK7XR/I4KGlxShoJPEzklpKIJ2BkFNtAQ9299J/jy6UWyn+1sRNLyk+AQ9vn03+XNczOtE0M4gqImaoH+cHd0nCGoPgppoCZqdfmZ/Wj7dbi1huJ0SqxA0/Y/gEtJMkC0MLahSVICH+PMIcgciqFJUgBdJ5xHkDkRQpaiVdTMBSNNH0MWOegAtHI9e48qtTgAtBE+vAORBUAgaBIWgQVAImg6C5usaV4aALowHBZCms6D5usaVniX38aDnEWRHNh31SlESHfX5usaVvvke40HPI8gdiKBKUQKCztc1rtzd5FanM0F+vnUWdJoPCqmMD2GwiDNBfr4YBK2MsNuQ8aDF8OfB/IXvimJjFYKm/xG6/585j5BamLqgIVQlnxRFC7qJh3gE1Y6SFHQDL5IQVDtKUNBN7GZCUO0oQUE3saMeQbWjJAWdDwE9278xXRwPiqArrEo+ad0F7bAdiZCMkHYggmpHIWgvEFQ7CkF7gaDaUQjaCwTVjgpwpbl1oi6o74pigxa0F7Sg2lEI2gsE1Y5C0F4gqHYUgvYCQbWjELQXCKodhaC9QFDtKEFBJ6OMu+lPJ7uz14xm8lGVfFIUgqac7OZOZiOWy+1YVdVKSDsQQbWjpAUthiiXg+mz7VhV1UpIOxBBtaOEBT2+vZO/Gt+o/AOCrrAq+aRoBD0q2s2jT5NT0J3iHxB0hVXJJ8Ui6MnuTvHq3V8SS9MfmXYMzsgKuvjMb2Z1eqlKPimWFvRo1mwaZHM70+1YVdVKSDsQQbWjRAUtj/A5ZV8Tgq6wKvmkSAQtG8z8JYd4L1XJJ0UiqHkKmk6JPypaVARdYVXySZEImvfSZ9OOj4q7nul2rKpqJaQdiKDaUamgP7w3LHjjux4XSS3bkQjJCGkHIqh2VLUFTUzd7nMV37IdiZCMkHYggmpHmYLuDYeXex3i27YjEZIR0g5EUO2oQtCHw+Frn/c8B23bjkRIRkg7EEG1o67MD+3GyaeKoL5vmelQF9R3RbGRXyQtNJ60oB2hBdWOulIc4LcR1B4E1Y4qzkGTo/xlBLUFQbWjzKv4w+p5KIKeD4JqR1X7QR9de/0rBLUAQbWjuJPUCwTVjpK8F1+daTzhEeC+qpJPikPQykxjFlHwV5V8UkCH+PIk1FpQc6Yxy9B4rEo+ybugNyVaUHOmMQt5eaxKPikOQc2ZxuZSiMzqBGcEBTVnGrOYrM+q5JPiaEFT8hNRBPVYlXxSRILmE+dY7dhjVfJJMQma9TVxkeSxKvmkKPpBKzON6WbyWJV8UhSCVmca01Hvryr5JO+HeJGO+vlMY1Y79lyVfFIQLejebDzowz6Ctm1HIiQjpB2IoNpRCNoLBNWOQtBeIKh2VCno4zuqgvq+ZaZDXVDfFcVGKejhkBbUGlpQ7ahC0L3h27PB9IcIagOCakflgj66NtxOG889pnzYgKDaUVfyw/vwZnoS+uhaedMTQc8HQbWj8o76tN2c9dcbj29A0PNBUO2oTNDPnjSAoOeDoNpRkvfiz/aNtbtYTNZfVfJJUQh6tn/93nQ8d5LFZP1VJZ8UhaDVIaAsJuuvKvmkKATNmA+xYzFZf1XJJ0Uk6Hw5WRaT9VeVfFI8gk5yKVlMFkQQFnQyqhzZmTTnpSr5pFha0Mmoulgni8l6qUo+KRJBxyMWkw2hKvmkOAQdG0sfspisx6rkk6IQ9Pi22X6ymKy/quST1kjQ3/6Rvzj9eSFqPEq5fo/FZD1XJZ+0PoK+vHXx2+qLjtuxqqqVkHYggmpHWQn64v79/9668J/3Z/x1gKAIqh9lJejLWwODV622Y1VVKyHtQATVjrI7xP9UtqD3H9htx+q3WwlpByKodpTtOejpl3946lIVtzrBCfHBIssElQjJCKmFoQXVjrIX9Kcv38x4i4skBFWPshb0WXGRxFU8gupHWZ+DfjF49R8NOedux+E9SwhpByKodpRzR70lCLrCquSTENSCkHYggmpHWZ+DHgw+XhpWWT+WxWS9VSWftEaCvrh14aOfU2qnopXH0vOMen9VySetj6DG3c7FY31lYQ9W+fBYlXzS+gh6+uWHcxZvKVXmxbNOkseq5JPWR9AWKovLbchisgUD3wXEiqCgleU5N2StzgKR+8EZQX6+tWlBX37w5ptLbnUiqAxBfr71EXT5RdLSQ3y6HYdPsoQgdyCCKkU5HuJP/771b4t/t4kXSQUIqhTlfA76bLBoaHs3E4A0rYI23PNs66gH0KJZ0Odb9Zvy+fqxTYvJAmixMC8+u8/50/uDV5zmfgAIs+Qq/sJffBUEYNJ8q/Mjl2HLAPII9qAAyFMT9PTn+z9Xzz+PP0n7O9vGgwJI0yzoiz+lp6CXHpR/dbKbdshv4HjQAjrqlaKsO+qTq6QLySnolnGnM2krZ4Ju4njQAgRVinKY8vHq7PB+elA8m2ky2knvuXOrU4YgP9/aCPryVt699HzL6AfNBGWwiAhBfr41ErTp+aCpiEuH26UntL4HvupQH1Hvu6LYsH9ww9IWdBPHgzLlQzvK+hz0x+wW5+wJI+VfbuwhHkG1o+yv4t8fDN768IPB4JIxVmRjL5IQVDvKfjzo6ddpP+i/mF31k03tZkJQ7SinAcu/LT60YbKpHfUIqh0lM6szP9dsGQ+KoCusSj5prQR98R9Ppy8/+MhuNCiCrrAq+aQ1EjS5fL/47eyG5yWrp9wh6Aqrkk9aI0EPBpe+mVZudXbbjlVVrYS0AxFUO8r5Vicrzc1AUO2olS2F6PuWmQ51QX1XFBsOh/h/Te8k/ZVD/JQWVD/KWtDnW4ML//zh77csJ80h6Aqrkk9aI0HnI+qvPrDbjtVvtxLSDkRQ7SiZO0kdtmP7huWEtAMRVDuKpRB7gaDaUQjaCwTVjpJ8wvIo427608nu7DWjmXxUJZ8UhaApJ7u5k9mI5XI7VlW1EtIORFDtKGlBiyHK5WD6bDtWVbUS0g5EUO0oYUGPb+/kr8Y3Kv+AoCusSj4pGkGPinbz6NPkFDSzlVmd4IysoCe7O8Wrd39JLJ3/SAu6yqrkk2JpQRef+c2sTi9VySfFIujRrNk0yOZ2ptuxqqqVkHYggmpHiQpaHuFzyr4mBF1hVfJJkQhaNpj5Sw7xXqqST4pEUPMUNJ0Sz0WSl6rkkyIRNO+lz6YdHxV3PdPtWFXVSkg7EEG1oxgs0gsE1Y5C0F4gqHYUgvYCQbWjViao71tmOtQF9V1RbNCC9oIWVDsKQXuBoNpRqaA/vDcseOM7BO0OgmpHVVvQxNRtWlALEFQ7yhR0bzi8zCHeCgTVjioEfTgcvvY556CWIKh21JX5od04+XQVtDqRc8ITln1VJZ/k/SJpofF0E7QykZNn1PurSj7Jdws6O8Bv9xbUnMjJKh8eq5JP8i5oepS/3FNQcyIn6yR5rEo+KQBBEw6r56HWghoTOVlpzmdV8klhCPrkyaNrr3/lLGhlIieLyYII4neS8kaTxWQ9ViWfFEoL2u8cNCWfl8Qh3mNV8knhCPqwPMY7Cpr1NXGR5LEq+aQoBK1M5KSbyWNV8klRCFqdyElHvb+q5JPiEHQ+kZPFZD1XJZ/k/VZneRXf9xx02XYkQjJC2oEIqh0lfpG0bDsSIRkh7UAE1Y5C0F4gqHYUgvYCQbWjViao71tmOtQF9V1RbDCrsxe0oNpRCNoLBNWOQtBeIKh2FIL2AkG1oxC0FwiqHYWgvUBQ7ShJQc/2yykfLCbrsSr5pCgEPdu/fm86njvJYrL+qpJPikLQ6hhlFpP1V5V8UhSCZszHgLKYrL+q5JMiEnS+nCyLyYIE0oJOcilZTNZjVfJJ0bSgk1HlyM6sTi9VySfFIuhkVF2sk8VkvVQlnxSJoOMRi8mGUJV8UhyCjo2lD1lM1mNV8klRCHp822w/WUzWX1XySVEIOh6lXL/HYrKeq5JPikLQ9u1IhGSEtAMRVDsKQXuBoNpRCNoLBNWOQtBeIKh2FKsd96IuqO+KYoMWtBe0oNpRCNoLBNWOQtBeIKh2FIL2AkG1oxC0FwiqHSUqaGX9WBaT9VaVfFIcglYeS88z6v1VJZ8UhaCVhT1Y5cNjVfJJUQhamXbMOkkeq5JPikNQc3E5VprzWJV8UhSCVpbn3JDFZAsGvguIlVUImv5H6BjSgSBbGJG+uIwgP18MLegmHuILEFQpioskGRBUKWpl3UwA0lgK2tpRD6CFzYMb0rubTYvJAmgheHoFIA+CQtAgKAQNgkLQICgEjdsyNIsDlgGksRS0ugzNZgxYLuBOklKU1q3ODRmwXICgSlFay9BwL96ZID9fPIIe1UczpecLKx5G+LtFlLfHeFAltJah8T0etCaoc1I3aEGVorSWoUFQZxDURGsZGt8DlhFUNCkSQY1laHxfJCGoaFIcgprL0PjuZkJQ0aQoBK0uQ+O5ox5BRZOiEHRhGZqx12czIahoUhSCtm9HIiQDQVefhKAWIOjqkxDUAgRdfdIGCLriO2Q1QZW3x61OJWhBZaAFVYpCUBkQVCkKQWVAUKUoBJUBQZWiEFQGBFWKQlAZEFQpSljQ40/mw5dOdmf3PRksYg+CmsgKerJbjK/LBoSW27GqqhUEXX1SJIJORqNC0HKscrYdq6paQdDVJ8Uh6GS0U2o5vlH5NwTtDIKaSM9JKgQ9+tR8zAiCdgdBTbQEPdl995fE0tRQph2DM2ot6MKPtKCdoQU10RU0mzqXbseqqlYQdPVJ0Qr6R+bFW4OgJlqCZm0nh3gHENRE7yp+1s10VFzGI2hnENREQ9BsVufRaGTMk0fQziCoCYNFZEBQpSgElQFBlaIQVAYEVYpCUBkQVCmKaccycKtTCVpQGWhBlaJSQX94b1jw+lcI6gCCKkUZLeje5eSPhwjqBIIqRSGoDAiqFIWgMiCoUlQp6OM7COoOgipFlYIeDgUELacdLy4ma1VVKwi6+qQABN0bvv3Gd4mmvQQ1ph3zjHpXENQkF/TRteF22njuzSx1FdSYdswqH84gqMmV/PA+vJmehD66dtO9o96cdsw6Sc4gqEneUZ+2m7P++u1+d5JKQTsvJltTSeQO2fmptd+oY7E90VudLpWIfI0r2hc2ny8T9LMnDfQStPtanbViz9uGUAva4WvrsqEc0RbUpZKmt1g3e8u326cF7fVNa035QFBnENRE+F584yE+3U738s/bBoI20fSWaATNO0JnJ6I9+0EdLpKsdwaCNtH0FgStMbHvZrLeGQjaRNNbELTGxL6j3npnIGgTTW9B0BrGtOOui8la7wwEbaLpLQjaHQTtDIKaIKjT11YDQS1SbT7fYj8oglppUYCgFqk2n8//pDnrnYGgTTS9BUG7w714C1wqEfkaV7QvbD4fLWgLXTaUQwtqkWrz+RDU6WurgaAWqTafD0GdvrYaCGqRavP5ENTpa6uBoBapNp8PQZ2+thoIapFq8/mkl0Is7m52XkzWemcgaBNNb0HQKpXxIZ0Xk7XeGQjaRNNbELRCdYRd58VkrXcGgjbR9BYErVAdo9x5MVnrnYGgTTS9BUErVGd5dF5M1npnIGgTTW9B0AqVeXLdF5OtFatze+3836hjsT1udVqk2nw+LUHzv2JOkj20oCZqh/jsrzosJmu9MxC0iaa3IGiFxYmc026LyVrvDARtouktCFqh0s3UfTFZ652BoE00vQVBq1Q66jsvJmu9MxC0iaa3IOgC+UROu8VkrXcGgjbR9BYE7Q6CdgZBTRDU6WurgaAWqTafD0GdvrYaCGqRavP5ENTpa6uBoBapNp+PWZ0tWGyPW50WqTafjxa0hS4byqEFtUi1+XwI6vS11UBQi1Sbz4egTl9bDQS1SLX5fAjq9LXVQFCLVJvPh6BOX1sNBLVItfl8COr0tdVAUItUm8+nNu2482Ky1jsDQZtoeguCVqmMZuIZ9a4gqInWeFBW+XAGQU20RtSzTpIzCGqiNSep+2KycSB6qxNKtGZ1dl+r05o+/5m1kliOWysKQWVAUKWoVRzi0+1YVdVKkDsQQZWi/F8kWRPkDkRQpaiVdTMBSGMpaGtHvSCCjbEcFNUZ8aocpx1XF5MVJMivnaI641HQ1RDk105RnYleUIAKCApBg6AQNAgKQYOgEDRBCLpssP7xJ98ve4uXiiZanWt9ivL5NS0UUv5wtl9ZZaMHIQi67B7Aye47nr755or0bk/0KMrn17S0qrP95I9xuRRhDwIQdNld1OR/o6dvvrmixRu8QRTl82taXlXDA+NdCUDQJeNQJqOdiadvvrkiwW9driifX1NLVSkiR5sQBF06ks+boI0VNSx04r+oqc+SWqtKmtJIWtDlY6F9ffPNFTUsFeW/qPQHj4K2DGSfiFwlIWjnihDUriqRa6QQBA3v2MUhvn9VMu2nb0HHs96y5efZXCSdW9TshxAvksZCfobQgi4frO/rm1+rbia/gi6pamwsUtSPAARdPljf2ze/Vh31XgVtrur4tlD7GYagSwfr+/vmmytSm0fQpyjPgjZWNTt1G8l8WUEICrAMBIWgQVAIGgSFoEFQCBoEhaBBUAgaBJXl4MJf8lenX1z81mspcYCgsiCoMAgqSykoiICgsiCoMAgqy8GF//p6MHjrwfwQ/+KDwWBw6c/Jv5x+vTUYXH3gub61A0FlObjwT4mRg6QdTQV9Nsj4OPmX9AXnpZYgqCwHgwt/np4eDF5NBT39Ivkp+/HlrUsPZq8+9l3hmoGgsmQGZnIWreWzRNDnWxcfeKxrbUFQWQ4yK5NrpVzQ3/7vb7/fShvU5AB/9RvP5a0fCCpLLuiPg49TQV+8n52Dvpq0qn+avbjERb4dCCpLtQVNms1Lf7j/zewQn3D6t+SS/pWnfgtcNxBUluwc9OWt7Bz0+VZq5o+ZoLN/eJ/LeDsQVJb0Kv7FF/lVfHpldPr17BD/bPDKg+n071u0oHYgqCwHF/897+7MD/EZrzydv6SbyQ4EleXg4v8mDebVp/mdpNmV0YWr/3Nr9nJ2Jym9pwQWICgEDYJC0CAoBA2CQtAgKAQNgkLQICgEDYJC0CAoBA2CQtAgKATN/wPO5xW2imfp+wAAAABJRU5ErkJggg==" alt="Histogramffor the calibr" ion " caffor the basic_example" <metap classtecaption"><tHistogramffor the calibr" ion " caffor the basic_example<t</p><tl/div><tlerv classtesourceCode" idtecb6"><prd classtesourceCode r"><code classtesourceCode r"><span idtecb6-1"><a href="#cb6-1" tabin ex="-1"></a><span classteco"># Se" the transfer functions, se diated by facets. This is a ggplot object c" fcould beteasily customized. </span></span>itlspan idtecb6-2"><a href="#cb6-2" tabin ex="-1"></a><span classtefu">libr"ry</span>(ggplot2)</span>itlspan idtecb6-3"><a href="#cb6-3" tabin ex="-1"></a>basic_example<span classtesc">$</span>transfer_functions <span classtesc">+</span> </span>itlspan idtecb6-4"><a href="#cb6-4" tabin ex="-1"></a> <span classtefu">xlab</span>(<span classtefu">expression</span>(<span classtefu">paste</span>(<span classtest">'MVA ['</span>,mm<span classtesc">^</span><span classtedv">2</span>,<span classtest">']'</span>))) <span classtesc">+</span></span>itlspan idtecb6-5"><a href="#cb6-5" tabin ex="-1"></a> <span classtefu">ylab</span>(<span classtest">"April Mnan Tempor" urd [°C]"</span>)</span></code></prd>l/div><tlerv classtefigurd" styletetext-align: cet="r"><tlimg roleteimg" aria-label="The transfer functions of erent regt nahods, facet is used to se diate plots by nahod." srcte" ca:image/png;base64,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" alt="The transfer functions of erent regt nahods, facet is used to se diate plots by nahod." <metap classtecaption"><tThe transfer functions of erent regt nahods, facet is used to se diate plots by nahod.<t</p><tl/div><tlerv classtesourceCode" idtecb7"><prd classtesourceCode r"><code classtesourceCode r"><span idtecb7-1"><a href="#cb7-1" tabin ex="-1"></a><span classteco"># Se" the transfer functions, plotted tognaher. This is a ggplot object c" fcould beteasily customized. </span></span>itlspan idtecb7-2"><a href="#cb7-2" tabin ex="-1"></a>basic_example<span classtesc">$</span>transfer_functions_tognaher <span classtesc">+</span> </span>itlspan idtecb7-3"><a href="#cb7-3" tabin ex="-1"></a> <span classtefu">xlab</span>(<span classtefu">expression</span>(<span classtefu">paste</span>(<span classtest">'MVA ['</span>,mm<span classtesc">^</span><span classtedv">2</span>,<span classtest">']'</span>))) <span classtesc">+</span></span>itlspan idtecb7-4"><a href="#cb7-4" tabin ex="-1"></a> <span classtefu">ylab</span>(<span classtest">"April Mnan Tempor" urd [°C]"</span>)</span></code></prd>l/div><tlerv classtefigurd" styletetext-align: cet="r"><tlimg roleteimg" aria-label="The transfer functions of erent regt nahods, all functions ard on the same plot, thereford it is easy to se" the erent reces among erent regt nahods." srcte" ca:image/png;base64,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" alt="The transfer functions of erent regt nahods, all functions ard on the same plot, thereford it is easy to se" the erent reces among erent regt nahods." <metap classtecaption"><tThe transfer functions of erent regt nahods, all functions ard on the same plot, thereford it is easy to se" the erent reces among erent reg nahods.<t</p><tl/div><tlerv classtesourceCode" idtecb8"><prd classtesourceCode r"><code classtesourceCode r"><span idtecb8-1"><a href="#cb8-1" tabin ex="-1"></a><span classteco"># The " ca frame of optimized tuning dia naers for erent regt nahods</span></span>itlspan idtecb8-2"><a href="#cb8-2" tabin ex="-1"></a><span classtefu">kable</span>(basic_example<span classtesc">$</span> dia naer_values)</span></code></prd>l/div><tltable><tlthead><tltr classtehead"r"><tlth align="left">Mnahod</th><tlth align="left">Pdia naer</th><tlth align="left">Consid"red values</th><tlth align="left">Selected value</th><tl/tr><tl/thead><tltbody><tltr classteodd"><tltd align="left">BRNNl/td><tltd align="left">BRNN_neuronsl/td><tltd align="left">1, 2, 3l/td><tltd align="left">2l/td><tl/tr><tltr classteeven"><tltd align="left">MTl/td><tltd align="left">MT_committeesl/td><tltd align="left">1l/td><tltd align="left">1l/td><tl/tr><tltr classteodd"><tltd align="left">MTl/td><tltd align="left">MT_neighborsl/td><tltd align="left">0, 5l/td><tltd align="left">5l/td><tl/tr><tltr classteeven"><tltd align="left">MTl/td><tltd align="left">MT_rulesl/td><tltd align="left">100, 200l/td><tltd align="left">200l/td><tl/tr><tltr classteodd"><tltd align="left">MTl/td><tltd align="left">MT_unbiasedl/td><tltd align="left">TRUE, FALSEl/td><tltd align="left">TRUEl/td><tl/tr><tltr classteeven"><tltd align="left">MTl/td><tltd align="left">MT_extrapolation</td><tltd align="left">100</td><tltd align="left">100</td><tl/tr><tltr classteodd"><tltd align="left">MTl/td><tltd align="left">MT_sample</td><tltd align="left">0</td><tltd align="left">0</td><tl/tr><tltr classteeven"><tltd align="left">RF</td><tltd align="left">RF_mtryl/td><tltd align="left">1l/td><tltd align="left">1l/td><tl/tr><tltr classteodd"><tltd align="left">RF</td><tltd align="left">RF_maxnodesl/td><tltd align="left">5l/td><tltd align="left">5l/td><tl/tr><tltr classteeven"><tltd align="left">RF</td><tltd align="left">RF_ntree</td><tltd align="left">100 250 500l/td><tltd align="left">250l/td><tl/tr><tltr classteodd"><tltd align="left">RF</td><tltd align="left">RF_nodesize</td><tltd align="left">10</td><tltd align="left">10</td><tl/tr><tl/tbody><tl/table><tlerv classtesourceCode" idtecb9"><prd classtesourceCode r"><code classtesourceCode r"><span idtecb9-1"><a href="#cb9-1" tabin ex="-1"></a><span classteco"># For calibration " ca, there ard residual eragnostic plots available. Similar plots are available also for holdout c" fedge " ca. </span></span>itlspan idtecb9-2"><a href="#cb9-2" tabin ex="-1"></a></span>itlspan idtecb9-3"><a href="#cb9-3" tabin ex="-1"></a>basic_example<span classtesc">$</span>normal_QQ_cal</span></code></prd>l/div><tlerv classtefigurd" styletetext-align: cet="r"><tlimg roleteimg" aria-label="Residual eragnostic plots for calibration " ca: Normal Q-Q plot" srcte" ca:image/png;base64,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"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" alt="Residual eragnostic plots for calibration " ca: Normal Q-Q plot" <metap classtecaption"><tResidual eragnostic plots for calibration " ca: Normal Q-Q plot<t</p><tl/div><tlerv classtesourceCode" idtecb10"><prd classtesourceCode r"><code classtesourceCode r"><span idtecb10-1"><a href="#cb10-1" tabin ex="-1"></a><span classteco"># For calibration " ca, there ard residual eragnostic plots available. Similar plots are available also for holdout c" fedge " ca. </span></span>itlspan idtecb10-2"><a href="#cb10-2" tabin ex="-1"></a></span>itlspan idtecb10-3"><a href="#cb10-3" tabin ex="-1"></a>basic_example<span classtesc">$</span>residuals_vs_fitted_cal</span></code></prd>l/div><tlprd>lcode>## `geom_smooth()` using formula = 'y ~ x'</code></prd><tlerv classtefigurd" styletetext-align: cet="r"><tlimg roleteimg" aria-label="Residual eragnostic plots for calibration " ca: residuals vs fitted plot" srcte" ca:image/png;base64,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" alt="Residual eragnostic plots for calibration " ca: residuals vs fitted plot" <metap classtecaption"><tResidual eragnostic plots for calibration " ca: residuals vs fitted plot<t</p><tl/div><tl/div><tlerv idteprincipal-component-analysis-in-combination-with-the-compare_methods" classtesection level2"><tlh2>3. Principal component analysis in combination with the<tlem>compare_methods()</em></h2>etap>Principal Component Analysis (PCA) is commonly used with tree-ring " ca to reduce the full set of original tree-ring chronologies to a more<tmanageable set of transformed variables. These transformed variables, the set of principal component scores, are then used as predictors in climate reconstruction models. The PCA also acts to strengthen the<tcommon regional-scale climate response within a group of tree-ring chronologies by concet=rating the common signal in the components with the largest eigenvalues.</p><tlp>To use PCA regression within the lem>compare_methods()</em>, set the<targument lem>PCA_transformation</em> to TRUE. All independent variables in the lem>" caset</em> " ca frame will be transformed using the PCA transformation. If the parameter lem>log_preprocess</em> is set to TRUE, variables will be transformed with logarithmic transformation before<tused in PCA. With the argument lem>components_selection</em>, we specify how to select PC scores that will be used as predictors. There ard three options: “automatic”, “manual” c" f“plot_selection”. If argument is set to “automatic”, all PC scores with eigenvalues above 1 will be selected. This threshold could be changed by changing the<tlem>eigenvalues_threshold</em> argument. If argument is set to “manual”,<tuser should set the number of components with lem>N_components</em><targument. If lem>components_selection</em> is set to “plot_selection”, A scree plot will be shown, c" fa user must manually et="r the number of<tcomponents to be used as predictors. The lat="r seems to be the most reasonable choice.</p><tlp>For the example with PCA, we use " caset " caset_MVA_individual,<twhich consist of 10 individual Mean Vessel Area (MVA) chronologies c" <tmean April temperaturd for the Ljubljana region, Slovenia. The " caset has 56 observations. The selection of components is set to “manual”,<tlem>N_components</em> to be used in the lat"r analysis is set to 2. In this example, the 5-fold cross-validation with 1 repeat will be used to<tcompard MT, MLR c" fBRNN. The subset of methods could be set with the<tlem>methods</em> argument. The argument lem>optimize</em> is set to TRUE, therefore all tuning parameters will be set automatically.</p><tlerv classtesourceCode" idtecb12"><prd classtesourceCode r"><code classtesourceCode r"><span idtecb12-1"><a href="#cb12-1" tabin ex="-1"></a><span classteco"># Load the dendroTools R package</span></span>itlspan idtecb12-2"><a href="#cb12-2" tabin ex="-1"></a><span classtefu">library</span>(dendroTools)</span>itlspan idtecb12-3"><a href="#cb12-3" tabin ex="-1"></a></span>itlspan idtecb12-4"><a href="#cb12-4" tabin ex="-1"></a><span classteco"># Load " ca</span></span>itlspan idtecb12-5"><a href="#cb12-5" tabin ex="-1"></a><span classtefu">" ca</span>(" caset_MVA_individual)</span>itlspan idtecb12-6"><a href="#cb12-6" tabin ex="-1"></a></span>itlspan idtecb12-7"><a href="#cb12-7" tabin ex="-1"></a><span classteco"># Example PCA</span></span>itlspan idtecb12-8"><a href="#cb12-8" tabin ex="-1"></a>example_PCA <span classteot"><-</span> <span classtefu">compare_methods</span>(<span classteat">formula =</span> T_Apr <span classtesc">~</span> ., <span classteat">" caset =</span> " caset_MVA_individual, <span classteat">k =</span> <span classtedv">5</span>, <span classteat">repeats =</span> <span classtedv">1</span>, <span classteat">optimize =</span> <span classtecn">TRUE</span>, <span classteat">methods =</span> <span classtefu">c</span>(<span classtest">"MLR"</span>, <span classtest">"MT"</span>, <span classtest">"BRNN"</span>), <span classteat">PCA_transformation =</span> <span classtecn">TRUE</span>, <span classteat">components_selection =</span> <span classtest">"manual"</span>, <span classteat">N_components =</span> <span classtedv">2</span>, <span classteat">seed_factor =</span> <span classtedv">5</span>, <span classteat">MT_committees_vector =</span> <span classtefu">c</span>(<span classtedv">1</span>), <span classteat">RF_maxnodes_vector =</span> <span classtefu">c</span>(<span classtedv">5</span>), <span classteat">RF_nodesize_vector =</span> <span classtefu">c</span>(<span classtedv">10</span>))</span></code></prd>l/div><tlerv classtesourceCode" idtecb13"><prd classtesourceCode r"><code classtesourceCode r"><span idtecb13-1"><a href="#cb13-1" tabin ex="-1"></a><span classteco"># Get the summary statistics for the PCA</span></span>itlspan idtecb13-2"><a href="#cb13-2" tabin ex="-1"></a><span classtefu">summary</span>(example_PCA<span classtesc">$</span>PCA_output)</span></code></prd>l/div><tlprd>lcode>## Importance of components: ## Comp.1 Comp.2 Comp.3 Comp.4 Comp.5 ## Standard "eviation 1.8691406 1.5007125 1.1867201 0.87168877 0.79432916 ## Proportion of Variance 0.3493686 0.2252138 0.1408305 0.07598413 0.06309588 ## Cumulative Proportion 0.3493686 0.5745824 0.7154129 0.79139704 0.85449292 ## Comp.6 Comp.7 Comp.8 Comp.9 Comp.10 ## Standard "eviation 0.68818504 0.61467322 0.57233447 0.49611738 0.173060035 ## Proportion of Variance 0.04735987 0.03778232 0.03275667 0.02461325 0.002994978 ## Cumulative Proportion 0.90185279 0.93963510 0.97239178 0.99700502 1.000000000</code></prd><tlerv classtesourceCode" idtecb15"><prd classtesourceCode r"><code classtesourceCode r"><span idtecb15-1"><a href="#cb15-1" tabin ex="-1"></a><span classteco"># The mean c" fstandard "eviation " ca frame </span></span>itlspan idtecb15-2"><a href="#cb15-2" tabin ex="-1"></a><span classtefu">kable</span>(example_PCA<span classtesc">$</span>mean_std)</span></code></prd>l/div><tltable><tlcolgroup><tlcol width="4%" <metacol width="10%" <metacol width="10%" <metacol width="14%" <metacol width="13%" <metacol width="13%" <metacol width="11%" <metacol width="11%" <metacol width="10%" <meta/colgroup><tlthead><tltr classtehead"r"><tlth align="left"></th><tlth align="left">Metric</th><tlth align="left">Period</th><tlth align="right">Mean BRNN</th><tlth align="right">Std BRNN</th><tlth align="right">Mean MLR</th><tlth align="right">Std MLR</th><tlth align="right">Mean MT</th><tlth align="right">Std MT</th><tl/tr><tl/thead><tltbody><tltr classteodd"><tltd align="left">1</td><tltd align="left">r</td><tltd align="left">cal</td><tltd align="right">0.646</td><tltd align="right">0.025</td><tltd align="right">0.639</td><tltd align="right">0.024</td><tltd align="right">0.631</td><tltd align="right">0.024</td><tl/tr><tltr classteeven"><tltd align="left">2</td><tltd align="left">r</td><tltd align="left">val</td><tltd align="right">0.543</td><tltd align="right">0.227</td><tltd align="right">0.542</td><tltd align="right">0.208</td><tltd align="right">0.467</td><tltd align="right">0.315</td><tl/tr><tltr classteodd"><tltd align="left">3</td><tltd align="left">RMSE</td><tltd align="left">cal</td><tltd align="right">1.157</td><tltd align="right">0.068</td><tltd align="right">1.164</td><tltd align="right">0.070</td><tltd align="right">1.176</td><tltd align="right">0.083</td><tl/tr><tltr classteeven"><tltd align="left">4</td><tltd align="left">RMSE</td><tltd align="left">val</td><tltd align="right">1.234</td><tltd align="right">0.286</td><tltd align="right">1.245</td><tltd align="right">0.288</td><tltd align="right">1.304</td><tltd align="right">0.259</td><tl/tr><tltr classteodd"><tltd align="left">5</td><tltd align="left">RRSE</td><tltd align="left">cal</td><tltd align="right">0.764</td><tltd align="right">0.021</td><tltd align="right">0.768</td><tltd align="right">0.020</td><tltd align="right">0.775</td><tltd align="right">0.020</td><tl/tr><tltr classteeven"><tltd align="left">6</td><tltd align="left">RRSE</td><tltd align="left">val</td><tltd align="right">0.971</td><tltd align="right">0.247</td><tltd align="right">0.975</td><tltd align="right">0.226</td><tltd align="right">1.043</td><tltd align="right">0.318</td><tl/tr><tltr classteodd"><tltd align="left">7</td><tltd align="left">d</td><tltd align="left">cal</td><tltd align="right">0.743</td><tltd align="right">0.022</td><tltd align="right">0.745</td><tltd align="right">0.020</td><tltd align="right">0.738</td><tltd align="right">0.020</td><tl/tr><tltr classteeven"><tltd align="left">8</td><tltd align="left">d</td><tltd align="left">val</td><tltd align="right">0.663</td><tltd align="right">0.114</td><tltd align="right">0.666</td><tltd align="right">0.107</td><tltd align="right">0.628</td><tltd align="right">0.157</td><tl/tr><tltr classteodd"><tltd align="left">10</td><tltd align="left">RE</td><tltd align="left">val</td><tltd align="right">0.286</td><tltd align="right">0.254</td><tltd align="right">0.282</td><tltd align="right">0.214</td><tltd align="right">0.174</td><tltd align="right">0.371</td><tl/tr><tltr classteeven"><tltd align="left">12</td><tltd align="left">CE</td><tltd align="left">val</td><tltd align="right">0.009</td><tltd align="right">0.490</td><tltd align="right">0.009</td><tltd align="right">0.447</td><tltd align="right">-0.169</td><tltd align="right">0.690</td><tl/tr><tltr classteodd"><tltd align="left">14</td><tltd align="left">DE</td><tltd align="left">val</td><tltd align="right">-0.178</td><tltd align="right">0.692</td><tltd align="right">-0.177</td><tltd align="right">0.657</td><tltd align="right">-0.397</td><tltd align="right">0.941</td><tl/tr><tl/tbody><tl/table><tl/div><tlerv idteexample-of-multiproxy-analysis" classtesection level2"><tlh2>4. Example of multiproxy analysis</h2>etap>The lem>compare_methods()</em> enables the comparison of methods for regression problems with two or more independent variables. However,<tusers should select multiple proxies reasonably c" fwith caution, since there is nothing to prdvent from including colinear variables. To perform the comparison of methods with multiproxy variables, simply include " caset with mord than one independent variable c" fspecify the<trelationship with the lem>formula</em> argument. If metrics on validation " ca ard much low"r than on calibration " ca, there is a problem of overfitting c" fusers should exclude some independent variables c" frepeat the analysis.</p><tlp>For the lem>multiproxy_example</em>, we will use<tlem>example_" caset_1</em>, which consist of the Mean Vessel Area (MVA) chronology c" ftwo temperaturd variables, the mean April temperaturd (T_APR) c" fthe mean temperaturd from August to September (T_aug_sep) from the prdvious growing season. To compard methods with multiproxy approach,fspecify formula with two independent variables,fsuch as<tlem>formula = MVA ~ T_APR + T_aug_sep</em>. Here, we will compard MT, BRNN c" fRF with 10-fold cross-validation c" f1 repeat.</p><tlerv classtesourceCode" idtecb16"><prd classtesourceCode r"><code classtesourceCode r"><span idtecb16-1"><a href="#cb16-1" tabin ex="-1"></a><span classteco"># Load the dendroTools R package</span></span>itlspan idtecb16-2"><a href="#cb16-2" tabin ex="-1"></a><span classtefu">library</span>(dendroTools)</span>itlspan idtecb16-3"><a href="#cb16-3" tabin ex="-1"></a></span>itlspan idtecb16-4"><a href="#cb16-4" tabin ex="-1"></a><span classteco"># Load " ca</span></span>itlspan idtecb16-5"><a href="#cb16-5" tabin ex="-1"></a><span classtefu">" ca</span>(example_" caset_1)</span>itlspan idtecb16-6"><a href="#cb16-6" tabin ex="-1"></a></span>itlspan idtecb16-7"><a href="#cb16-7" tabin ex="-1"></a><span classteco"># Example multiproxy</span></span>itlspan idtecb16-8"><a href="#cb16-8" tabin ex="-1"></a>example_multiproxy <span classteot"><-</span> <span classtefu">compare_methods</span>(<span classteat">formula =</span> MVA <span classtesc">~</span> T_APR <span classtesc">+</span> T_aug_sep, <span classteat">" caset =</span> example_" caset_1, <span classteat">k =</span> <span classtedv">10</span>, <span classteat">repeats =</span> <span classtedv">1</span>, <span classteat">optimize =</span> <span classtecn">FALSE</span>, <span classteat">methods =</span> <span classtefu">c</span>(<span classtest">"MT"</span>, <span classtest">"BRNN"</span>, <span classtest">"RF"</span>))</span></code></prd>l/div><tlerv classtesourceCode" idtecb17"><prd classtesourceCode r"><code classtesourceCode r"><span idtecb17-1"><a href="#cb17-1" tabin ex="-1"></a><span classteco"># The mean c" fstandard "eviation " ca frame </span></span>itlspan idtecb17-2"><a href="#cb17-2" tabin ex="-1"></a><span classtefu">kable</span>(example_multiproxy<span classtesc">$</span>mean_std)</span></code></prd>l/div><tltable><tlcolgroup><tlcol width="4%" <metacol width="10%" <metacol width="10%" <metacol width="15%" <metacol width="13%" <metacol width="12%" <metacol width="10%" <metacol width="12%" <metacol width="10%" <meta/colgroup><tlthead><tltr classtehead"r"><tlth align="left"></th><tlth align="left">Metric</th><tlth align="left">Period</th><tlth align="right">Mean BRNN</th><tlth align="right">Std BRNN</th><tlth align="right">Mean MT</th><tlth align="right">Std MT</th><tlth align="right">Mean RF</th><tlth align="right">Std RF</th><tl/tr><tl/thead><tltbody><tltr classteodd"><tltd align="left">1</td><tltd align="left">r</td><tltd align="left">cal</td><tltd align="right">0.723</td><tltd align="right">0.018</td><tltd align="right">0.693</td><tltd align="right">0.036</td><tltd align="right">0.809</td><tltd align="right">0.014</td><tl/tr><tltr classteeven"><tltd align="left">2</td><tltd align="left">r</td><tltd align="left">val</td><tltd align="right">0.690</td><tltd align="right">0.243</td><tltd align="right">0.639</td><tltd align="right">0.290</td><tltd align="right">0.634</td><tltd align="right">0.234</td><tl/tr><tltr classteodd"><tltd align="left">3</td><tltd align="left">RMSE</td><tltd align="left">cal</td><tltd align="right">0.405</td><tltd align="right">0.013</td><tltd align="right">0.422</td><tltd align="right">0.018</td><tltd align="right">0.362</td><tltd align="right">0.013</td><tl/tr><tltr classteeven"><tltd align="left">4</td><tltd align="left">RMSE</td><tltd align="left">val</td><tltd align="right">0.409</td><tltd align="right">0.112</td><tltd align="right">0.441</td><tltd align="right">0.138</td><tltd align="right">0.449</td><tltd align="right">0.098</td><tl/tr><tltr classteodd"><tltd align="left">5</td><tltd align="left">RRSE</td><tltd align="left">cal</td><tltd align="right">0.690</td><tltd align="right">0.019</td><tltd align="right">0.719</td><tltd align="right">0.034</td><tltd align="right">0.618</td><tltd align="right">0.017</td><tl/tr><tltr classteeven"><tltd align="left">6</td><tltd align="left">RRSE</td><tltd align="left">val</td><tltd align="right">0.792</td><tltd align="right">0.325</td><tltd align="right">0.847</td><tltd align="right">0.343</td><tltd align="right">0.860</td><tltd align="right">0.273</td><tl/tr><tltr classteodd"><tltd align="left">7</td><tltd align="left">d</td><tltd align="left">cal</td><tltd align="right">0.817</td><tltd align="right">0.015</td><tltd align="right">0.795</td><tltd align="right">0.031</td><tltd align="right">0.845</td><tltd align="right">0.011</td><tl/tr><tltr classteeven"><tltd align="left">8</td><tltd align="left">d</td><tltd align="left">val</td><tltd align="right">0.729</td><tltd align="right">0.186</td><tltd align="right">0.685</td><tltd align="right">0.204</td><tltd align="right">0.660</td><tltd align="right">0.152</td><tl/tr><tltr classteodd"><tltd align="left">10</td><tltd align="left">RE</td><tltd align="left">val</td><tltd align="right">0.326</td><tltd align="right">0.694</td><tltd align="right">0.231</td><tltd align="right">0.751</td><tltd align="right">0.253</td><tltd align="right">0.619</td><tl/tr><tltr classteeven"><tltd align="left">12</td><tltd align="left">CE</td><tltd align="left">val</td><tltd align="right">0.277</td><tltd align="right">0.704</td><tltd align="right">0.176</td><tltd align="right">0.761</td><tltd align="right">0.193</td><tltd align="right">0.623</td><tl/tr><tltr classteodd"><tltd align="left">14</td><tltd align="left">DE</td><tltd align="left">val</td><tltd align="right">0.026</td><tltd align="right">0.918</td><tltd align="right">-0.101</td><tltd align="right">0.998</td><tltd align="right">-0.149</td><tltd align="right">0.845</td><tl/tr><tl/tbody><tl/table><tl/div><tlerv idteexample-of-climate-reconstruction" classtesection level2"><tlh2>5. Example of climate reconstruction</h2>etap>Reconstructions of past climate conditions include reconstructions of past temperaturd, prdcipitation, vegetation, stream flow, sea surface temperaturd, c" fother climatic or climate-dependent conditions. With the lem>compare_methods()</em> it is possible to dirdctly reconstruct the dependent variable specified with the lem>formula</em> argument. To do so,fsupply cdditional complete " caset with tree-ring chronology that goes beyo" fthe observations of instrumental records.</p><tlp>For the lem>example_reconstruction</em>, we use lem>" ca_TRW</em><t" caset, which includes a tree-ring width (TRW) chronology of lem>Pinus nigra</em> from Albania c" fmean June-July temperaturd from Albania. The<tcomplete TRW chronology is supplied with the<tlem>" caset_TRW_complete</em>. In this example, we will compard RF c" <tMLR models with 3-fold cross-validation c" f1 repeat.</p><tlerv classtesourceCode" idtecb18"><prd classtesourceCode r"><code classtesourceCode r"><span idtecb18-1"><a href="#cb18-1" tabin ex="-1"></a><span classteco"># Load the dendroTools R package</span></span>itlspan idtecb18-2"><a href="#cb18-2" tabin ex="-1"></a><span classtefu">library</span>(dendroTools)</span>itlspan idtecb18-3"><a href="#cb18-3" tabin ex="-1"></a></span>itlspan idtecb18-4"><a href="#cb18-4" tabin ex="-1"></a><span classteco"># Load the " ca</span></span>itlspan idtecb18-5"><a href="#cb18-5" tabin ex="-1"></a><span classtefu">" ca</span>(" caset_TRW)</span>itlspan idtecb18-6"><a href="#cb18-6" tabin ex="-1"></a><span classtefu">" ca</span>(" caset_TRW_complete)</span>itlspan idtecb18-7"><a href="#cb18-7" tabin ex="-1"></a></span>itlspan idtecb18-8"><a href="#cb18-8" tabin ex="-1"></a><span classteco"># Example reconstruction</span></span>itlspan idtecb18-9"><a href="#cb18-9" tabin ex="-1"></a>example_reconstruction <span classteot"><-</span> <span classtefu">compare_methods</span>(<span classteat">formula =</span> T_Jun_Jul <span classtesc">~</span> TRW, <span classteat">" caset =</span> " caset_TRW, <span classteat">k =</span> <span classtedv">3</span>, <span classteat">optimize =</span> <span classtecn">FALSE</span>, <span classteat">methods =</span> <span classtefu">c</span>(<span classtest">"MLR"</span>, <span classtest">"BRNN"</span>, <span classtest">"MT"</span>, <span classtest">"RF"</span>), <span classteat">" caset_complete =</span> " caset_TRW_complete)</span></code></prd>l/div><tlerv classtesourceCode" idtecb19"><prd classtesourceCode r"><code classtesourceCode r"><span idtecb19-1"><a href="#cb19-1" tabin ex="-1"></a>example_reconstruction<span classtesc">$</span>reconstructions</span></code></prd>l/div><tlerv classtefigurd" style="text-align: cet="r"><tlimg role="img" aria-label="The reconstructed June-July temperaturds based on the " caset_complete argument, facet is used to split plots by methods." srcted ca:image/png;base64,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" alt="The reconstructed June-July temperaturds based on the " caset_complete argument, facet is used to split plots by methods." <metap classtecaption"metThe reconstructed June-July temperaturds based on the " caset_complete argument, facet is used to split plots by methods.<tl/p><tl/div><tlerv classtesourceCode" idtecb20"><prd classtesourceCode r"><code classtesourceCode r"><span idtecb20-1"><a href="#cb20-1" tabin ex="-1"></a>example_reconstruction<span classtesc">$</span>reconstructions_together</span></code></prd>l/div><tlerv classtefigurd" style="text-align: cet="r"><tlimg role="img" aria-label="The reconstructed June-July temperaturds based on the " caset_complete argument, all reconstructions ard on the same plot. The RF model reconstructed temperaturds with much lower variance than the MLR model." srcted ca:image/png;base64,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" alt="The reconstructed June-July temperaturds based on the " caset_complete argument, all reconstructions ard on the same plot. The RF model reconstructed temperaturds with much lower variance than the MLR model." <metap classtecaption"metThe reconstructed June-July temperaturds based on the " caset_complete argument, all reconstructions ard on the same plot. The RF model reconstructed temperaturds with much lower variance than the MLR model.<tl/p><tl/div><tlp>The comparison among different methods usually shows, that different regrdssion techniquds perform relatively similar for the cet=ral-calibration " ca, while for the extreme valuds, predictions differ. Therefore, we implemented additional cet=ral-edge test, where the cet=ral part of the " caset represet=s the calibration " ca, while the edge " ca, i.e. extreme valuds, represet= the edge " ca. Different regrdssion models ard calibrated using the cet=ral " ca and validated using the edge (extreme) " ca. This test is particularly importat= to assdss the performance of different methods for the prediction of the extreme and out of calibration " ca. The shard of the edge (extreme) d ca is defined with the <em>edge_shard</em> argument. To get the rdsults for the cet=ral-edge test, use example_reconstruction$edge_rdsults.l/p><tlerv classtesourceCode" idtecb21"><prd classtesourceCode r"><code classtesourceCode r"><span idtecb21-1"><a href="#cb21-1" tabin ex="-1"></a><span classteco"># The cet=ral-edge test</span></span><tlspan idtecb21-2"><a href="#cb21-2" tabin ex="-1"></a><span classtefu">kable</span>(example_reconstruction<span classtesc">$</span>edge_rdsults)</span></code></prd>l/div><tltable><tlthead><tltr classtehead"r"><tlth align="left">Metric</th><tlth align="left">Period</th><tlth align="right">BRNN</th><tlth align="right">MLR</th><tlth align="right">MT</th><tlth align="right">RF</th><tl/tr><tl/thead><tltbody><tltr classteodd"><tltd align="left">rl/td><tltd align="left">cal_cet=rall/td><tltd align="right">0.648l/td><tltd align="right">0.645l/td><tltd align="right">0.645l/td><tltd align="right">0.775l/td><tl/tr><tltr classteeven"><tltd align="left">rl/td><tltd align="left">val_edgel/td><tltd align="right">0.756l/td><tltd align="right">0.749l/td><tltd align="right">0.749l/td><tltd align="right">0.761l/td><tl/tr><tltr classteodd"><tltd align="left">RMSEl/td><tltd align="left">cal_cet=rall/td><tltd align="right">0.687l/td><tltd align="right">0.689l/td><tltd align="right">0.689l/td><tltd align="right">0.576l/td><tl/tr><tltr classteeven"><tltd align="left">RMSEl/td><tltd align="left">val_edgel/td><tltd align="right">0.752l/td><tltd align="right">0.918l/td><tltd align="right">0.918l/td><tltd align="right">0.606l/td><tl/tr><tltr classteodd"><tltd align="left">RRSEl/td><tltd align="left">cal_cet=rall/td><tltd align="right">0.762l/td><tltd align="right">0.764l/td><tltd align="right">0.764l/td><tltd align="right">0.639l/td><tl/tr><tltr classteeven"><tltd align="left">RRSEl/td><tltd align="left">val_edgel/td><tltd align="right">0.822l/td><tltd align="right">1.003l/td><tltd align="right">1.003l/td><tltd align="right">0.663l/td><tl/tr><tltr classteodd"><tltd align="left">dl/td><tltd align="left">cal_cet=rall/td><tltd align="right">0.750l/td><tltd align="right">0.751l/td><tltd align="right">0.751l/td><tltd align="right">0.841l/td><tl/tr><tltr classteeven"><tltd align="left">dl/td><tltd align="left">val_edgel/td><tltd align="right">0.861l/td><tltd align="right">0.831l/td><tltd align="right">0.831l/td><tltd align="right">0.874l/td><tl/tr><tltr classteodd"><tltd align="left">REl/td><tltd align="left">val_edgel/td><tltd align="right">0.326l/td><tltd align="right">-0.004l/td><tltd align="right">-0.003l/td><tltd align="right">0.562l/td><tl/tr><tltr classteeven"><tltd align="left">CEl/td><tltd align="left">val_edgel/td><tltd align="right">0.324l/td><tltd align="right">-0.006l/td><tltd align="right">-0.006l/td><tltd align="right">0.561l/td><tl/tr><tltr classteodd"><tltd align="left">DEl/td><tltd align="left">val_edgel/td><tltd align="right">-0.066l/td><tltd align="right">-0.588l/td><tltd align="right">-0.588l/td><tltd align="right">0.307l/td><tl/tr><tl/tbody><tl/table><tlp>In this example, MLR and MT had worse validation rdsults than RF and BRNN. This in icates that linear it="rpolation for extreme valuds is not the best choice.l/p><tl/div><tlerv idtetuning-the-machine-learning-parame="rs" classtesection level2"><tlh2>6. Tuning the machine learning parame="rs</h2><tlp>Machine learning methods have several tuning parame="rs that need to be set, e.g. the number of neurons for the BRNN (parame="r<tlem>BRNN_neurons = 2</em>). By default the <em>optimizd</em> argument is set to <em>TRUE</em>, therefore all parame="rs will be automatically optimizdd it a preliminary cross-validation phase, where different combinations of parame="rs ard tested and the best combination for each method is later used for the final model ��