Facet plot to show flows for each year. See the blog https://wordpress.com/post/tonyladson.wordpress.com/10393

facet.R

######################################################################################
#
# Facet plot for flows
# 
#######################################################################################

library(tidyverse)
library(lubridate)
library(gridExtra)
library(scales)
library(padr)


# Read in sample flow data
# Available on Google Drive at
# https://drive.google.com/open?id=1-g1rOfeOxquPow4YM1aifdEkdBboUzu4

# For advice on reading in files stored on Google Drive
# https://stackoverflow.com/questions/33135060/read-csv-file-hosted-on-google-drive

id <- "1-g1rOfeOxquPow4YM1aifdEkdBboUzu4" # google file ID

flow_daily <- read_csv(file = sprintf("https://docs.google.com/uc?id=%s&export=download", id),
                       col_types = cols(
                         X1 = col_character(),
                         X2 = col_double(),
                         X3 = col_integer(),
                         X4 = col_double(),
                         X5 = col_integer(),
                         X6 = col_double(),
                         X7 = col_integer()
                       ),
                       col_names = FALSE,
                       skip = 17)


# On my computer
# project_path <- "/Users/anthonyladson/Dropbox/Grad Cert/2017/Unit 1/Hydrology/Casey's Weir/"
# flow_daily <- read_csv(file =str_c(project_path, '404216_20170628/404216.csv' ),
#                        col_types = cols(
#                          X1 = col_character(),
#                          X2 = col_double(),
#                          X3 = col_integer(),
#                          X4 = col_double(),
#                          X5 = col_integer(),
#                          X6 = col_double(),
#                          X7 = col_integer()
#                        ),
#                        col_names = FALSE,
#                        skip = 17
# )

names(flow_daily) <- c('date',
                       'flow_mean',
                       'qcode_mean',
                       'flow_min',
                       'qcode_min',
                       'flow_max',
                       'qcode_max')

# Convert date

flow_daily <- flow_daily %>% 
  mutate(date = as.Date(str_sub(date, 1, 10), format = '%d/%m/%Y'))  # Convert time, just use days


#______________________________________________________________________________________
# Checks


# start and end of the file

head(flow_daily)
tail(flow_daily)


# Missing data

colSums(is.na(flow_daily))

# Remove first line (contains missing data and add columns we'll use later)

flow_daily <- flow_daily %>% 
  slice(2:nrow(flow_daily)) # There is only one missing value, which is the first, so lets remove it.


# Check time step

x <- flow_daily %>% 
  mutate(date_diff = date - lag(date)) %>% 
  count(date_diff) # ok
res <- assertthat::assert_that(x[1,2] == nrow(flow_daily)-1) 



# Zero values

colSums(dplyr::near(flow_daily[,-1], 0))
# looks fine
# date  flow_mean qcode_mean   flow_min  qcode_min   flow_max  qcode_max  flow_year flow_month 
# 0        560          0        783          0        557          0          0          0 
# jday 
# 0 


# plot
flow_daily %>% 
  ggplot(aes(date, flow_mean)) + 
  geom_line() +
  scale_x_date(name = 'Date',
               date_breaks = '2 years',
               date_labels = '%Y',
               limits = c(as.Date('1972-06-01'), as.Date('2017-7-31'))) + # These limits are a work around to so x-axis goes from 1972-2018
  scale_y_continuous(name = 'Mean daily flow (ML/d)', 
                     labels = comma) +
  theme(axis.text.x = element_text(angle = 45, hjust = 1))


# Facet plot

# Add additional date information
flow_daily <- flow_daily %>% 
  mutate(flow_year = year(date)) %>% 
  mutate(flow_month = month(date)) %>% 
  mutate(jday = yday(date))



month_starts <- yday(as.Date('2018-01-01') %m+% months(c(0, 3, 6, 9)))
month_starts
month_labs <- month.abb[c(1,4,7,10)]
month_labs

# with constant y-axis scaling

flow_daily %>% 
  ggplot(aes(jday, flow_mean)) +
  geom_line() +
  scale_y_continuous(name = 'Mean daily flow (ML/d)',
                     labels = comma) +
  scale_x_continuous(name = 'Date',
                     breaks = month_starts,
                     labels = month_labs) +
  facet_wrap(~flow_year, ncol = 5, scales = 'fixed')


# with variable y-axis scaling

flow_daily %>% 
  ggplot(aes(jday, flow_mean)) +
  geom_line() +
  scale_y_continuous(name = 'Mean daily flow (ML/d)',
                     labels = comma) +
  scale_x_continuous(name = 'Date',
                     breaks = month_starts,
                     labels = month_labs) +
  facet_wrap(~flow_year, ncol = 5, scales = 'free_y')