v1.2, refined selection plots in header

01_Extract_data_Project_31.Rmd

@@ -16,9 +16,10 @@ always_allow_html: yes
         
 **Timestamp:** `r date()`  
 **Drafted:** Francesco Maria Sabatini  
-**Version:** 1.1 
-  
+**Version:** 1.2  
 
+*Changes to version 1.2* - Corrected final selection of plots in header. Only plots with CWM info are retained, but no minimum trait coverage threshold is used.  
+<br>
 This report documents the data extraction for **sPlot project proposal #31** - *The adaptive value of xylem physiology within and across global ecoregions* as requested by Daniel Laughlin and Jesse Robert Fleri 
 
 
@@ -180,7 +181,7 @@ DT.xylem %>%
 
 
 # 3 Calculate CWMs and trait coverage 
-Calculate CWM and trait coverage for each trait and each plot. Select plots having more than 80% coverage for at least one trait.
+Calculate CWM and trait coverage for each trait and each plot. ~~Select plots having more than 80% coverage for at least one trait.~~
 ```{r, cache=T,  cache.lazy=F}
 # Merge species data table with traits
 CWM.xylem0 <- DT.xylem %>%
@@ -264,34 +265,43 @@ knitr::kable(CWM.xylem[1:20,], caption="Example of CWM data file") %>%
     kable_styling(bootstrap_options = c("striped", "hover", "condensed", "responsive"), full_width = F, position = "center")
 ```
 
-```{r}
-CWM.xylem08 <- CWM.xylem %>%
-  filter(trait.coverage>=0.8)
+```{r fig.width=8, fig.height=5, dpi=150, fig.align="center", fig.cap="Density of plots with CWM info, for increasing trait coverage."}
+ggtc <- ggplot(data=CWM.xylem) +
+  geom_density(aes(trait.coverage, col=trait)) + 
+  xlab("Trait coverage (abundance)") +
+  theme_bw()
+
+ggnsp <- ggplot(data=CWM.xylem) +
+  geom_density(aes(trait.coverage.nspecies, col=trait)) + 
+  xlab("Trait coverage (Presence absence)") +
+  theme_bw()
+
+cowplot::plot_grid(ggtc, ggnsp, nrow=1)
+
+
 ```
 
-```{r, echo=F}
-knitr::kable(CWM.xylem08 %>%
-               group_by(trait) %>%
-               summarize("num.plots"=n(), .groups="drop"), caption="Number of plots with >=.8 coverage per trait") %>%
-    kable_styling(bootstrap_options = c("striped", "hover", "condensed", "responsive"), full_width = F, position = "center")
+```{r, echo=F, eval=F}
+#knitr::kable(CWM.xylem08 %>%
+#               group_by(trait) %>%
+#               summarize("num.plots"=n(), .groups="drop"), caption="Number of plots with >=.8 coverage per trait") %>%
+#    kable_styling(bootstrap_options = c("striped", "hover", "condensed", "responsive"), full_width = F, position = "center")
 ```
 
 
 ```{r}
-#Create list of plots having at least one trait with >=.8 coverage and extract header data
+#Create list of plots for which CWM data is available
+plot.sel2 <- CWM.xylem %>%
+      distinct(PlotObservationID) %>% 
+      pull(PlotObservationID)
 
-#plot80perc <- (CWM.xylem %>%
-#  dplyr::select(PlotObservationID) %>%
-#  distinct())$PlotObservationID
+header.xylem <- header.xylem %>% 
+  filter(PlotObservationID %in% plot.sel2)
 
-#DT.xylem08 <- DT.xylem %>%
-#  filter(PlotObservationID %in% header.xylem$PlotID)
-
-#CWM.xylem <- CWM.xylem %>%
-#  filter(PlotObservationID %in% header.xylem$PlotID) 
 ```
 
-Completeness of header data
+
+Evaluate completeness of header data
 ```{r, echo=F}
 knitr::kable(data.frame(Completeness_perc=colSums(!is.na(header.xylem))/
                           nrow(header.xylem)*100)[-c(1,2),,drop=F], 
@@ -333,7 +343,7 @@ The data derive from `r nrow(data.origin)` datasets.
 
 ```{r}
 soilclim.xylem <- soilclim %>% 
-  filter(PlotObservationID %in% plot.sel) %>% 
+  filter(PlotObservationID %in% plot.sel2) %>% 
   dplyr::rename(Elevation=Elevation_median) %>% 
   dplyr::select(-Elevation_q2.5, -Elevation_q97.5, -Elevation_DEM.res)
 ```