nanosims_analysis_fractionation.Rmd

---
title: "NanoSIMS Data Analysis - Fractionation"
author: "A. Pasulka, S. Kopf"
date: "`r Sys.Date()`"
output: html_document
---

```{r "load libraries", echo = FALSE, message=FALSE, warning=FALSE}
library(tidyverse)
library(knitr)
opts_chunk$set(dev=c("png", "pdf"), dev.args=list(pdf = list(encoding="WinAnsi", useDingbats=FALSE)),
  fig.keep="all", fig.path = file.path("plot", "NANOSIMS_fractionation_"))
```

## Load data
```{r}
load(file.path("data", "nanosims_calculations.RData"))
```

## Visualize fractionation factors

Estimated from combined ROIs' total fractionation vs. circumnatural ratios expected for unlabelled samples. 

```{r, fractionation_factors_by_run_FINAL, fig.width=10, fig.height=8, warning=FALSE}
nanosims_calculations %>% 
  filter(grepl("alpha", data_type)) %>% 
  select(RunDate, variable, value, sigma) %>% 
  mutate(variable = sub("alpha ", "", variable)) %>% 
  distinct() %>% 
  mutate(inferred = ifelse(RunDate %in% c("June_2015", "Sept_2015") & 
                             variable == "15N", "approximated from all runs", 
                           "measured in run")) %>% 
  ggplot() + 
  aes(x = RunDate, y = value - 1, fill = RunDate, alpha = inferred) +
  geom_bar(position = "dodge", stat = "identity", color = "black") +
  geom_errorbar(position = "dodge", size = 0.2, 
                aes(ymin = value - sigma - 1, ymax = value + sigma - 1)) + 
  geom_hline(yintercept = 0, size = 0.5) +
  facet_grid(variable~., scales = "free_y") + 
  scale_y_continuous(expression("Fractionation "~epsilon['NS/expected']),
                     labels = function(x) paste(1000*x, "\U2030")) +
  scale_alpha_manual(name = "", values = c(0.4, 1)) +
  scale_fill_brewer(palette = "Set1") +
  theme_bw() +
  theme(axis.text.x = element_text(angle = 90, vjust = 0.5, hjust=1))
```