finished 2b run models, ready to run

2a_process_survey.Rmd

@@ -81,7 +81,7 @@ dat$TOTAL <- HAUL$vertebrate_weight_kg
 dat$TOTAL[which(is.na(dat$TOTAL))] <- 0 # replace NA with 0
 ```
 
-```{r}
+```{r message=FALSE}
 # Add catch of each species by haul (takes a couple min)
 library(dplyr)
 dat$YEYE <- dat$DBRK <- dat$PHLB <- 0
@@ -94,14 +94,21 @@ for(i in 1:n.hauls){
 }
 # Order by date
 dat <- dat[order(dat$DATE),]
+# Add DAY covariate: day of the year
+dat$DAY <- as.numeric(dat$DATE - as.Date(paste0(dat$YEAR,"-01-01")))
+# Add binomial catch columns
+dat$DBRK_01 <- dat$PHLB_01 <- dat$YEYE_01 <- 0
+dat$DBRK_01[which(dat$DBRK>0)] <- 1
+dat$PHLB_01[which(dat$PHLB>0)] <- 1
+dat$YEYE_01[which(dat$YEYE>0)] <- 1
 ```
 
 ### Add SST covariate
   - `SST`: daily sea surface temperature anomalies (in degC)
 
 Download daily sea surface temperature anomalies (.nc files) for 2003-2013 from: https://www.esrl.noaa.gov/psd/data/gridded/data.noaa.oisst.v2.highres.html 
 
-```{r eval=TRUE, echo=TRUE, results='hide'}
+```{r eval=TRUE, echo=TRUE, message=FALSE, results='hide'}
 # function to get SST daily anomaly at the DATE/LON/LAT for each haul
 #   uses bilinear interpolation from nearest gridpoints
 library(ncdf4)
@@ -164,7 +171,7 @@ dat <- dat[-which(is.na(dat$SST)),]
 
 *Note:* The Rockfish Conservation Area (RCA) boundaries have changed by month, year, latitude, and depth. We have prepared `rca_boundaries.csv` using historical RCA boundaries. For more details, see the [RCA webpage](http://www.westcoast.fisheries.noaa.gov/fisheries/management/groundfish_closures/rockfish_areas.html).
 
-```{r}
+```{r message=FALSE}
 library(tidyr)
 
 # Get historical RCA boundary limits
@@ -217,12 +224,19 @@ dat$logDEPTH <- log(dat$DEPTH)
 # Center/de-mean each covariate
 dat$sst <- dat$SST
 demean <- function(vec){ return(vec - mean(vec))}
-dat[,c("logDEPTH","sst")] <- apply(dat[,c("logDEPTH","sst")],2,demean)
+dat[,c("DAY","logDEPTH","sst")] <- apply(dat[,c("DAY","logDEPTH","sst")],2,demean)
 
 # Create squared covariates
 dat$sst2 <- dat$sst^2
 dat$logDEPTH2 <- dat$logDEPTH^2
 
+# Turn categorical variables into factors
+dat$YEAR <- as.factor(dat$YEAR)
+dat$DBRK_01 <- as.factor(dat$DBRK_01)
+dat$PHLB_01 <- as.factor(dat$PHLB_01)
+dat$YEYE_01 <- as.factor(dat$YEYE_01)
+dat$inRCA <- as.factor(dat$inRCA)
+
 # Data are ready to fit
 save(dat, file="/home/brian/Documents/Bycatch/WCGOP/data/wcann_processed.RData")
 ```

2a_process_survey.html

@@ -134,16 +134,8 @@ <h3>Combine <code>HAUL</code> and <code>CATCH</code></h3>
 dat$TOTAL &lt;-<span class="st"> </span>HAUL$vertebrate_weight_kg
 dat$TOTAL[<span class="kw">which</span>(<span class="kw">is.na</span>(dat$TOTAL))] &lt;-<span class="st"> </span><span class="dv">0</span> <span class="co"># replace NA with 0</span></code></pre></div>
 <div class="sourceCode"><pre class="sourceCode r"><code class="sourceCode r"><span class="co"># Add catch of each species by haul (takes a couple min)</span>
-<span class="kw">library</span>(dplyr)</code></pre></div>
-<pre><code>## 
-## Attaching package: 'dplyr'</code></pre>
-<pre><code>## The following objects are masked from 'package:stats':
-## 
-##     filter, lag</code></pre>
-<pre><code>## The following objects are masked from 'package:base':
-## 
-##     intersect, setdiff, setequal, union</code></pre>
-<div class="sourceCode"><pre class="sourceCode r"><code class="sourceCode r">dat$YEYE &lt;-<span class="st"> </span>dat$DBRK &lt;-<span class="st"> </span>dat$PHLB &lt;-<span class="st"> </span><span class="dv">0</span>
+<span class="kw">library</span>(dplyr)
+dat$YEYE &lt;-<span class="st"> </span>dat$DBRK &lt;-<span class="st"> </span>dat$PHLB &lt;-<span class="st"> </span><span class="dv">0</span>
 for(i in <span class="dv">1</span>:n.hauls){
   <span class="co"># get all species caught in the ith haul</span>
     cur_haul &lt;-<span class="st"> </span><span class="kw">filter</span>(CATCH, trawl_id==dat$HAUL_ID[i])
@@ -152,7 +144,14 @@ <h3>Combine <code>HAUL</code> and <code>CATCH</code></h3>
     if(<span class="st">&quot;Pacific halibut&quot;</span> %in%<span class="st"> </span>cur_haul$common_name) dat$PHLB[i] &lt;-<span class="st"> </span><span class="kw">as.numeric</span>(dplyr::<span class="kw">filter</span>(cur_haul,common_name==<span class="st">&quot;Pacific halibut&quot;</span>) %&gt;%<span class="st"> </span>dplyr::<span class="kw">select</span>(total_catch_wt_kg))
 }
 <span class="co"># Order by date</span>
-dat &lt;-<span class="st"> </span>dat[<span class="kw">order</span>(dat$DATE),]</code></pre></div>
+dat &lt;-<span class="st"> </span>dat[<span class="kw">order</span>(dat$DATE),]
+<span class="co"># Add DAY covariate: day of the year</span>
+dat$DAY &lt;-<span class="st"> </span><span class="kw">as.numeric</span>(dat$DATE -<span class="st"> </span><span class="kw">as.Date</span>(<span class="kw">paste0</span>(dat$YEAR,<span class="st">&quot;-01-01&quot;</span>)))
+<span class="co"># Add binomial catch columns</span>
+dat$DBRK_01 &lt;-<span class="st"> </span>dat$PHLB_01 &lt;-<span class="st"> </span>dat$YEYE_01 &lt;-<span class="st"> </span><span class="dv">0</span>
+dat$DBRK_01[<span class="kw">which</span>(dat$DBRK&gt;<span class="dv">0</span>)] &lt;-<span class="st"> </span><span class="dv">1</span>
+dat$PHLB_01[<span class="kw">which</span>(dat$PHLB&gt;<span class="dv">0</span>)] &lt;-<span class="st"> </span><span class="dv">1</span>
+dat$YEYE_01[<span class="kw">which</span>(dat$YEYE&gt;<span class="dv">0</span>)] &lt;-<span class="st"> </span><span class="dv">1</span></code></pre></div>
 </div>
 <div id="add-sst-covariate" class="section level3">
 <h3>Add SST covariate</h3>
@@ -273,12 +272,19 @@ <h3>Standardize covariates</h3>
 <span class="co"># Center/de-mean each covariate</span>
 dat$sst &lt;-<span class="st"> </span>dat$SST
 demean &lt;-<span class="st"> </span>function(vec){ <span class="kw">return</span>(vec -<span class="st"> </span><span class="kw">mean</span>(vec))}
-dat[,<span class="kw">c</span>(<span class="st">&quot;logDEPTH&quot;</span>,<span class="st">&quot;sst&quot;</span>)] &lt;-<span class="st"> </span><span class="kw">apply</span>(dat[,<span class="kw">c</span>(<span class="st">&quot;logDEPTH&quot;</span>,<span class="st">&quot;sst&quot;</span>)],<span class="dv">2</span>,demean)
+dat[,<span class="kw">c</span>(<span class="st">&quot;DAY&quot;</span>,<span class="st">&quot;logDEPTH&quot;</span>,<span class="st">&quot;sst&quot;</span>)] &lt;-<span class="st"> </span><span class="kw">apply</span>(dat[,<span class="kw">c</span>(<span class="st">&quot;DAY&quot;</span>,<span class="st">&quot;logDEPTH&quot;</span>,<span class="st">&quot;sst&quot;</span>)],<span class="dv">2</span>,demean)
 
 <span class="co"># Create squared covariates</span>
 dat$sst2 &lt;-<span class="st"> </span>dat$sst^<span class="dv">2</span>
 dat$logDEPTH2 &lt;-<span class="st"> </span>dat$logDEPTH^<span class="dv">2</span>
 
+<span class="co"># Turn categorical variables into factors</span>
+dat$YEAR &lt;-<span class="st"> </span><span class="kw">as.factor</span>(dat$YEAR)
+dat$DBRK_01 &lt;-<span class="st"> </span><span class="kw">as.factor</span>(dat$DBRK_01)
+dat$PHLB_01 &lt;-<span class="st"> </span><span class="kw">as.factor</span>(dat$PHLB_01)
+dat$YEYE_01 &lt;-<span class="st"> </span><span class="kw">as.factor</span>(dat$YEYE_01)
+dat$inRCA &lt;-<span class="st"> </span><span class="kw">as.factor</span>(dat$inRCA)
+
 <span class="co"># Data are ready to fit</span>
 <span class="kw">save</span>(dat, <span class="dt">file=</span><span class="st">&quot;/home/brian/Documents/Bycatch/WCGOP/data/wcann_processed.RData&quot;</span>)</code></pre></div>
 </div>