In fishvice/pax: tidy fish data wrangling functions

knitr::opts_chunk$set(echo = TRUE, eval = FALSE)

To do

Get rid of repetition

library(dplyrOracle)
library(mar)
db <- src_oracle("mar")
Station <- 
  husky::STODVAR %>% 
  bind_rows() %>% 
  filter(tognumer %in% 1:39) %>% 
  select(id = synis.id,
         year = ar,
         towlength = toglengd,
         strata = newstrata,
         reitur,
         tognumer,
         lon,
         lat) %>% 
  mutate(towlength = pax:::trim_towlength(towlength),
         mult = 1) %>% 
  arrange(id) %>% 
  mutate(sid = paste(reitur, tognumer)) %>% 
  select(-c(reitur, tognumer))

# Use same strata for 2017 stations as last year
x <- 
  Station %>% 
  filter(year == 2016) %>% 
  select(sid, strata)

st <-
  lesa_stodvar(db) %>% 
  mutate(lon = ifelse(!is.na(hift_v_lengd), (kastad_v_lengd + hift_v_lengd)/2, kastad_v_lengd),
         lat = ifelse(!is.na(hift_n_breidd), (kastad_n_breidd + hift_n_breidd)/2, kastad_n_breidd)) %>% 
  filter(veidarfaeri %in % 73,
         ar == 2017,
         tognumer %in% 1:39) %>% 
  select(id = synis_id,
         year = ar,
         towlength = toglengd,
         reitur,
         tognumer,
         lon,
         lat) %>% 
  collect(n = Inf) %>% 
  mutate(towlength = pax:::trim_towlength(towlength),
         mult = 1,
         sid = paste(reitur, tognumer)) %>% 
  select(-c(reitur, tognumer)) %>% 
  left_join(x) %>% 
  filter(!is.na(strata))

table(!is.na(st$strata))

st %>% 
  ggplot(aes(lon, lat)) +
  geom_point() +
  coord_quickmap()

Station <-
  Station %>% 
  filter(sid %in% st$sid) %>% 
  bind_rows(st) %>% 
  select(id, year, towlength, strata, lon, lat) %>% 
  mutate(mult = 1)

Stratas <-
  husky::stratas_df %>% 
  select(strata, area = rall.area)

res <- calc_length_indices(Station, Stratas, SPECIES = 1, oracle = TRUE)

tidy_fixed <-
  res$aggr %>% 
  filter(length == 5) %>% 
  mutate(source = "tidy")
tidy_fixed %>% 
  select(year, cb, cb.cv) %>% 
  ggplot(aes(year, cb)) +
  geom_pointrange(aes(ymin = cb * (1 - cb.cv),
                      ymax = cb * (1 + cb.cv)),
                  colour = "red", lwd = 1) +
  geom_line(colour = "red", lwd = 1) +
  scale_colour_brewer(palette = "Set1")

res$aggr %>% 
  filter(year %in% c(2005:2017)) %>% 
  ggplot(aes(length, n)) +
  geom_line() +
  facet_wrap(~ year)

x <-
  res$aggr %>% 
  filter(year %in% c(2014:2017)) %>% 
  group_by(length) %>% 
  mutate(n.mean = mean(n),
         b.mean = mean(b))
x %>% 
  ggplot(aes(length)) +
  theme_bw() +
  geom_polygon(aes(y = n.mean), fill = "grey") +
  geom_ribbon(aes(ymin = n * (1 - n.cv),
                  ymax = n * (1 + n.cv)), 
              fill = "pink",
              alpha = 0.6) +
  geom_line(aes(y = n), colour = "red") +
  facet_grid(year ~ .) +
  labs(x = NULL, y = NULL,
       title = "Abundance indices",
       subtitle = "All years: only tows that have so far been taken in 2017") 

x %>% 
  ggplot(aes(length)) +
  theme_bw() +
  geom_polygon(aes(y = b.mean), fill = "grey") +
  geom_ribbon(aes(ymin = b * (1 - b.cv),
                  ymax = b * (1 + b.cv)), 
              fill = "pink",
              alpha = 0.6) +
  geom_line(aes(y = b), colour = "red") +
  facet_grid(year ~ .) 

cpue <- 
  catch_by_station(1, Station, std = "towlength", oracle = TRUE) 
cpue %>% 
  filter(length == 5,
         year %in% 2014:2017) %>% 
  ggplot(aes(lon, lat, size = cb)) +
  theme_bw() +
  geom_point(col = "red", alpha = 0.4) +
  facet_wrap(~ year) +
  scale_size_area(max_size = 15)