workflow_demo.Rmd

---
title: "Data Processing Example"
author: "ROBIN Network"
date: "`r Sys.Date()`"
output: html_document
---

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE)
source("./main_functions.R")
```

# Aim for pipeline
Insert text explaining the main steps of the ROBIN pipeline.

 > Data format (columns: Agency, Station, Day, Flow, Flag)
 > Check for Level-1/Level-2 status based on completeness.
 > Summary statistics
  + Monthly thresholds
  + POT series
  + 30-day/10-day/7-day rolling totals
  + Annual/monthly/quarterly quantiles/mean
  + Annual 1-day/10-day/30-day maxima
  + Annual 7-day/30-day minima
 > Tests
  + Pettitt breakpoint test
  + Seasonal Mann-Kendall
  + Rolling Mann-Kendall-Snyers
 > Indicators
  + SSI
  + Intermittence
  + Mean Annual Flow (Standardised)
  + Aridity Index
  

# Initial setup
```{r read_in}
dailyflow_path <- "./41022_gdf.csv"
dailyflow_raw <- read.csv(dailyflow_path, skip=20, col.names=c("DAY","FLOW","FLAG"), fill=T, header=F)
dailyflow_raw$DAY <- lubridate::ymd(dailyflow_raw$DAY)
#dailyflow_raw <- read_in_robin(dailyflow_path)  # or
#dailyflow_raw <- read_in_robin_db(agency, station)
```

```{r check_for_level}
datum <- dailyflow_raw
datum <-dailyflow_raw %>% 
  dplyr::mutate(year = lubridate::year(DAY)) %>%
  group_by(year) %>%
  dplyr::summarise(count=n()) %>% 
  dplyr::mutate(
    enough = count > ifelse(year%%4 == 0 & year%%400 !=0, 366, 365)*0.95)



# Check for
any(rle(X)$lengths[rle(X)$values=="m"] > 365*2)

```