Reach01.Rmd

---
title: "Reach Scale Inputs"
output: html_notebook
---

# Inputs

## Load packages

```{r pkgLoad}
suppressMessages(library(tidyverse))
library(readxl)
library(lubridate)
```

## Read in

```{r}
r0 <- read_csv("./origData/Reach1.csv",
               na = c("", "-9999.0"))
problems()
```

Really only need to parse out the date from the `Date` field.

```{r}
r0
```

```{r}

r0 %>% 
  distinct(Comments)
```

Ah so we have BQL issues.

```{r}

r0 %>% 
  distinct(`...15`)
```

So column 15 is blank?

```{r}
glimpse(r0)
```



So `Sampling Time` is OK as a time variable. Can we parse out the date from `Date`?


How many formats?

```{r}
r0 %>% 
  distinct(Date) %>% 
  print(n = 42)
```

These all appear to be dmy hms, but better check AUS dates

```{r}
r0 %>% 
  filter(Biome == "AUS") %>% 
  distinct(Date)
```

That looks OK. if we push this through Excel we get many problems with multiple data formats in one column. Even fancy stuff [like this](https://stackoverflow.com/questions/13764514/how-to-change-multiple-date-formats-in-same-column) does not get us out of that bother. So... maybe the parent data base exports its csv with some set of attributes that Excel tries to parse, but inconsistently.

## Parsing out sample dates

```{r}
r1 <- r0 %>% 
  # filter(Biome == "AUS") %>% 
  # distinct(Date) %>% 
  separate(Date, into = c("dt", "tm", "am"), sep = " ") %>% 
  mutate(SampDate = mdy(dt)) %>% 
  select(Biome:ReachType, SampDate, 
         SampNum = `Sample number`,
         SampTime = `Sampling Time (hh:mm)`,
         Amm = `Ammonium (ug N/L)`,
         Cl = `Chloride tracer (mg/L)`,
         Br = `Bromide tracer (mg/L)`,
         Comm = Comments) %>% 
  mutate(Biome = factor(Biome),
         SiteNumber = factor(SiteNumber),
         ReachType = factor(ReachType))
```

## Data layout

All `SiteType` are "E"

```{r}
r1 %>% distinct(SiteType)
```


```{r}
xtabs(~ Biome + SiteNumber + ReachType, r1)
```


# Prelim plot

```{r}
ggplot(r1, aes(SampTime, Amm, colour = SiteNumber)) +
  facet_grid(Biome ~ ReachType, scales = "free_y") + 
  geom_line()

```