data-raw/cricket.R

In Sayani07/gravitas: Explore Probability Distributions for Bivariate Temporal Granularities

library(readr)
library(tsibble)
library(dplyr)

# cricket <- read_csv("data-raw/deliveries.csv")


deliveries <- read_csv("data-raw/deliveries_all.csv")
matches <- read_csv("data-raw/matches_all.csv")

cricket_season <- deliveries %>%
  dplyr::left_join(matches, by = c("match_id" = "id")) %>%
  dplyr::mutate(winner_team = if_else(winner == batting_team, "batting_team", "bowling_team"))


cricket_winner <- cricket_season %>%
  select(match_id, winner_team) %>%
  unique()
# making it uniform
# each over needs to have 6 balls

cricket_fltr_ball <- cricket_season %>%
  filter(wide_runs + noball_runs == 0) %>% # filtering out scores that are gained by no balls or wide
  # still have to filter out overs for which number of balls per over less than 6
  group_by(
    batting_team,
    match_id,
    inning,
    over
  ) %>%
  summarise(n_over = length(ball)) %>%
  filter(n_over == 6) %>%
  dplyr::mutate(key = paste(batting_team,
    match_id,
    inning,
    over,
    sep = "_"
  ))


cricket_over_crctd <- cricket_season %>%
  dplyr::mutate(cricket_key = paste(batting_team, match_id, inning, over, sep = "_")) %>%
  filter(cricket_key %in% cricket_fltr_ball$key) %>%
  filter(wide_runs + noball_runs == 0) %>%
  dplyr::mutate(ball_per_over = purrr::rep_along(match_id, 1:6))

# each innings need to have 20 overs

cricket_fltr_over <- cricket_over_crctd %>%
  filter(inning %in% c(1, 2)) %>%
  group_by(inning, match_id) %>%
  summarise(n = length(over)) %>%
  filter(n == 120) %>%
  dplyr::mutate(key = paste(match_id, inning, sep = "_"))

cricket_over_inning_crctd <- cricket_over_crctd %>%
  dplyr::mutate(over_key = paste(match_id, inning, sep = "_")) %>%
  filter(over_key %in% cricket_fltr_over$key) %>%
  filter(inning %in% c(1, 2))


# each match needs to have 2 innings

cricket_filtr_inning <- cricket_over_inning_crctd %>%
  group_by(match_id) %>%
  summarise(length(inning)) %>%
  filter(`length(inning)` == 240)

cricketdata <- cricket_over_inning_crctd %>%
  filter(match_id %in% cricket_filtr_inning$match_id) %>%
  filter(wide_runs + noball_runs == 0) %>%
  dplyr::mutate(ball_per_over = purrr::rep_along(match_id, 1:6)) %>%
  filter(inning %in% c(1, 2))

#
# cricket <- cricketdata %>% select(season, match_id,
#                                   inning, batting_team,
#                                   bowling_team, ball_per_over,
#                                   over, winner, dismissal_kind,
#                                   total_runs)

# cricket <- cricketdata %>% group_by(season,
#                                     match_id,
#                                     batting_team,
#                                     bowling_team,
#                                     inning,
#                                     over) %>%
#   summarise(runs_per_over = sum(total_runs),
#             run_rate = sum(total_runs)/length(total_runs)) %>%
#   ungroup()


# cricket_tsibble_all <- cricket_per_over %>%
#   ungroup() %>%
#   mutate(data_index = row_number()) %>%
#   as_tsibble(index = data_index)


cricket <- cricketdata %>%
  dplyr::mutate(
    wicket = if_else(is.na(dismissal_kind), 0, 1),
    dot_balls = if_else(total_runs == 0, 1, 0)
  ) %>%
  group_by(
    season, match_id, batting_team,
    bowling_team, inning, over
  ) %>%
  summarise(
    wicket = sum(wicket),
    dot_balls = sum(dot_balls),
    runs_per_over = sum(total_runs),
    run_rate = round(sum(total_runs) / length(total_runs))
  ) %>%
  dplyr::ungroup()


cricket <- cricket %>%
  dplyr::left_join(cricket_winner, by = c("match_id")) %>%
  select(
    season, match_id, batting_team,
    bowling_team,
    inning,
    over,
    wicket,
    dot_balls,
    runs_per_over,
    run_rate,
    winner_team
  )



# save(cricket, file = "data/cricket.rda", compress = "gzip", version = 2)


usethis::use_data(cricket, overwrite = TRUE, compress = "xz")