library(tidyverse)
library(tidyr)
library(stringr)
library(ggplot2)

corridor <-
  read.csv("~/Documents/IdeaProjects/matsim-duesseldorf/scenarios/input/all_dusseldorf_non_motorway_links.csv") %>%
  tibble()

baseline <-
  read.delim("baseline/001.450.linkstats.txt.gz") %>%
  tibble()

bikelane <-
  read_delim("~/Documents/IdeaProjects/matsim-duesseldorf/output/experiments-pieter/all_duesseldorf_non_motorway_decongestion/duesseldorf-25pct-dc_114-no-lanes-noMc-decongestion.100.linkstats.txt.gz") %>%
  tibble()


data <-
  corridor %>%
  select(ID) %>%
  left_join(baseline ,
            by = c("ID" = "LINK")) %>%
  select(-ORIG_ID,-FROM,-TO) %>%
  tidyr::pivot_longer(!ID:CAPACITY) %>%
  mutate(src = "baseline") %>%
  rbind(
    corridor %>%
      select(ID) %>%
      left_join(bikelane ,
                by = c("ID" = "LINK")) %>%
      select(-ORIG_ID,-FROM,-TO) %>%
      tidyr::pivot_longer(!ID:CAPACITY, ) %>%
      mutate(src = "decongested")
  ) %>%
  filter(!name %>% str_detect("(0)[\\.\\-]*(24)")) %>%
  mutate(
    type = str_to_lower(name),
    type = ifelse(name %>% str_detect("HRS"), "volume", type),
    type = ifelse(name %>% str_detect("TRAVEL"), "traveltime", type),
    hour = name %>% str_extract("[0-9\\.]+") %>% as.integer(),
    property = ifelse(name %>% str_detect("min"), "min", NA),
    property = ifelse(name %>% str_detect("max"), "max", property),
    property = ifelse(name %>% str_detect("avg"), "avg", property)
  ) %>%
  select(-name)


data %>%
  filter(type == "volume") %>%
  group_by(src,  hour, property) %>%
  summarise(value = sum(4 * value * LENGTH / CAPACITY) / sum(LENGTH)) %>%
  pivot_wider(names_from = property) %>%
  ggplot() +
  geom_ribbon(aes(
    x = hour,
    ymin = min,
    ymax = max,
    fill = src,
    alpha = I(0.5)
  )) +
  geom_line(aes(x = hour, y = avg, color = src)) +
  ylab("Length-weighted average Volume/Capacity ratio across Duesseldorf area")

data %>%
  filter(type == "volume") %>%
  group_by(src,  hour, property) %>%
  summarise(value = sum(4 * value * LENGTH / CAPACITY) / sum(LENGTH)) %>%
  ungroup() %>%
  pivot_wider(names_from = property) %>%
  select(source = src, hour, avg) %>%
  pivot_wider(names_from = source, values_from = avg) %>%
  write.csv("corridor_volume_lineplot.csv",
            row.names = F,
            quote = F)

data %>%
  filter(type == "volume") %>%
  group_by(src,  hour, property) %>%
  summarise(value = max(value * 4 / CAPACITY)) %>%
  # summarise(value = sum(value * LENGTH / CAPACITY) / sum(LENGTH)) %>%
  pivot_wider(names_from = property) %>%
  ggplot() +
  geom_ribbon(aes(
    x = hour,
    ymin = min,
    ymax = max,
    fill = src,
    alpha = I(0.5)
  )) +
  geom_line(aes(x = hour, y = avg, color = src)) +
  ylab("Maximum Volume/Capacity ratio encountered across Duesseldorf area")

data %>%
  filter(type == "volume") %>%
  group_by(src,  hour, property) %>%
  summarise(value = max(value * 4 / CAPACITY)) %>%
  # summarise(value = sum(value * LENGTH / CAPACITY) / sum(LENGTH)) %>%
  pivot_wider(names_from = property) %>%
  select(source = src, hour, avg) %>%
  pivot_wider(names_from = source, values_from = avg) %>%
  write.csv("corridor_maxvolume_lineplot.csv",
            row.names = F,
            quote = F)


data %>%
  filter(type == "traveltime") %>%
  group_by(src,  hour, property) %>%
  summarise(value = value %>% sum) %>%
  mutate(value = value / 60) %>%
  pivot_wider(names_from = property) %>%
  ggplot() +
  geom_ribbon(aes(
    x = hour,
    ymin = min,
    ymax = max,
    fill = src,
    alpha = I(0.5)
  )) +
  geom_line(aes(x = hour, y = avg, color = src)) +
  ylab("Total Duesseldorf interior link travel time (minutes)")

data %>%
  filter(type == "traveltime") %>%
  group_by(src,  hour, property) %>%
  summarise(value = value %>% sum) %>%
  mutate(value = value / 60) %>%
  pivot_wider(names_from = c(property, src)) %>%
  write.csv("corridor_traveltime_lineplot.csv",
            row.names = F,
            quote = F)


data %>%
  filter(type == "traveltime") %>%
  group_by(src,  hour, property) %>%
  summarise(
    value = value %>% sum,
    LENGTH = LENGTH %>% sum,
    FREESPEED = FREESPEED %>% mean
  ) %>%
  mutate(fstt = LENGTH / FREESPEED,
         value = value / fstt) %>%
  pivot_wider(names_from = property) %>%
  ggplot() +
  geom_ribbon(aes(
    x = hour,
    ymin = min,
    ymax = max,
    fill = src,
    alpha = I(0.5)
  )) +
  geom_line(aes(x = hour, y = avg, color = src)) +
  ylab("Total corridor round-trip travel time / fstt")


networkdata <-
  baseline  %>%
  select(-ORIG_ID,-FROM,-TO) %>%
  filter(!LINK %>% str_detect("pt")) %>%
  select(LINK, CAPACITY, names(baseline)[names(baseline) %>% str_detect("HRS[0-9\\.\\-]*avg")]) %>%
  tidyr::pivot_longer(!LINK, values_to = "baseline") %>%
  filter(!name %>% str_detect("(0)[\\.\\-]*(24)")) %>%
  mutate(
    type = str_to_lower(name),
    type = ifelse(name %>% str_detect("HRS"), "volume", type),
    type = ifelse(name %>% str_detect("TRAVEL"), "traveltime", type),
    newcol = name %>% str_extract("[0-9]{1,2}"),
    newcol = ifelse(type == "capacity", "capacity", newcol)
  ) %>%
  select(-name,-type) %>%
  left_join(
    bikelane %>%
      select(-ORIG_ID,-FROM,-TO) %>%
      filter(!LINK %>% str_detect("pt")) %>%
      select(LINK, CAPACITY, (1:length(bikelane))[names(bikelane) %>% str_detect("HRS[0-9\\.\\-]*avg")]) %>%
      tidyr::pivot_longer(!LINK, values_to = "decongested") %>%
      filter(!name %>% str_detect("(0)[\\.\\-]*(24)")) %>%
      mutate(
        type = str_to_lower(name),
        type = ifelse(name %>% str_detect("HRS"), "volume", type),
        type = ifelse(name %>% str_detect("TRAVEL"), "traveltime", type),
        newcol = name %>% str_extract("[0-9]{1,2}"),
        newcol = ifelse(type == "capacity", "capacity", newcol)
      ) %>%
      select(-name,-type)
  )

network_out <-
  networkdata %>%
  filter(newcol != "capacity") %>%
  group_by(LINK) %>%
  summarise(baseline = sum(baseline),
            decongested = sum(decongested)) #%>%
# head(25*2) %>%
# mutate(delta = bikelane - baseline)
# pivot_wider(id_cols=link, names_from = newcol, values_from = c(delta))

write.csv(
  network_out %>% select(id = LINK, volume = baseline),
  "baseline_volume.csv",
  row.names = F,
  quote = F
)

write.csv(
  network_out %>% select(id = LINK, volume = decongested),
  "decongested_volume.csv",
  row.names = F,
  quote = F
)
network_out %>% semi_join(corridor, by = c("LINK" = "ID")) %>% write.csv("corridor-only.csv", row.names =
                                                                           F, quote = F)

networkdata.tt <-
  baseline  %>%
  select(-ORIG_ID,-FROM,-TO) %>%
  filter(!LINK %>% str_detect("pt")) %>%
  select(LINK, CAPACITY, names(.)[names(.) %>% str_detect("TRAVELTIME[0-9\\.\\-]*avg")]) %>%
  tidyr::pivot_longer(!LINK, values_to = "baseline") %>%
  filter(!name %>% str_detect("(0)[\\.\\-]*(24)")) %>%
  mutate(
    type = str_to_lower(name),
    type = ifelse(name %>% str_detect("HRS"), "volume", type),
    type = ifelse(name %>% str_detect("TRAVEL"), "traveltime", type),
    newcol = name %>% str_extract("[0-9]{1,2}"),
    newcol = ifelse(type == "capacity", "capacity", newcol)
  ) %>%
  select(-name,-type) %>%
  left_join(
    bikelane %>%
      select(-ORIG_ID,-FROM,-TO) %>%
      filter(!LINK %>% str_detect("pt")) %>%
      select(LINK, (1:ncol(.))[names(.) %>% str_detect("TRAVELTIME[0-9\\.\\-]*avg")]) %>%
      tidyr::pivot_longer(!LINK, values_to = "bikelane") %>%
      filter(!name %>% str_detect("(0)[\\.\\-]*(24)")) %>%
      mutate(
        type = str_to_lower(name),
        type = ifelse(name %>% str_detect("HRS"), "volume", type),
        type = ifelse(name %>% str_detect("TRAVEL"), "traveltime", type),
        newcol = name %>% str_extract("[0-9]{1,2}"),
        newcol = ifelse(type == "capacity", "capacity", newcol)
      ) %>%
      select(-name,-type)
  )

network_out <-
  networkdata.tt %>%
  filter(newcol != "capacity") %>%
  group_by(LINK) %>%
  summarise(baseline = mean(baseline),
            bikelane = mean(bikelane)) %>%
  # head(25*2) %>%
  left_join(baseline %>%
              mutate(fstt = LENGTH / FREESPEED) %>%
              select(LINK, fstt)) %>%
  mutate(
    baseline = 100 * baseline / fstt,
    bikelane = 100 * bikelane / fstt,
    delta = bikelane - baseline
  )
# pivot_wider(id_cols=link, names_from = newcol, values_from = c(delta))

write.csv(network_out,
          "corridor-traveltime.csv",
          row.names = F,
          quote = F)
network_out %>% semi_join(corridor, by = c("LINK" = "ID")) %>% write.csv("corridor-only-traveltime.csv",
                                                                         row.names = F,
                                                                         quote = F)



write.csv(
  network_out %>% select(id = LINK, ttratio = baseline),
  "baseline_tt.csv",
  row.names = F,
  quote = F
)

write.csv(
  network_out %>%
    select(id = LINK, ttratio = bikelane) %>%
    left_join(corridor %>%
                transmute (id =
                             ID, corridor = "corr")) %>% 
    mutate(corridor = ifelse(is.na(corridor),"no",corridor)
           ),
  "bikelane_tt.csv",
  row.names = F,
  quote = F
)