The below cells read in a .txt file, clean data, perform some calculations and display results as a dataframe. The data set used is information on a chess tournament that includes name of player, location, rank before and after the tournament, and information on each match.
Load data into rows and and create two lists of information. The first contains the id, name, points, and round information. The second line contains the state, pre-rank, and post-rank information. By selecting only every third row the borders are a non factor.
raw_text <- readLines(con = '/Users/kevinpotter/Documents/spring_2020_ms/data_607/project_1/tournamentinfo.txt')
line_1 <- raw_text[seq(5, length(raw_text), by = 3)]
line_1[1]## [1] " 1 | GARY HUA |6.0 |W 39|W 21|W 18|W 14|W 7|D 12|D 4|"line_2 <- raw_text[seq(6, length(raw_text), by = 3)]
line_2[1]## [1] " ON | 15445895 / R: 1794 ->1817 |N:2 |W |B |W |B |W |B |W |"Split the information in line_1 and line_2 into individual long vectors on the pipe “|” character. Then pull out the player_name, total_points, and state into separate lists using the seq package. Convert the total_points vector to a float and combine the three vectors into one dataframe. Get rid of the whitespace after the player_id using trimws.
vect_1 <- unlist(strsplit(line_1, '[|]'))
vect_2 <- unlist(strsplit(line_2, '[|]'))
player_name <- vect_1[seq(2, length(vect_1), by = 10)]
player_name <- trimws(player_name, which = c("right"), whitespace = "[ \t\r\n]")
total_points <- as.numeric(vect_1[seq(3, length(vect_1), by = 10)])
state <- vect_2[seq(1, length(vect_2), by = 10)]
df <- data.frame(player_name, state, total_points)
head(df)## player_name state total_points
## 1 GARY HUA ON 6.0
## 2 DAKSHESH DARURI MI 6.0
## 3 ADITYA BAJAJ MI 6.0
## 4 PATRICK H SCHILLING MI 5.5
## 5 HANSHI ZUO MI 5.5
## 6 HANSEN SONG OH 5.0Create a column for the rank of the player prior to the competition and state. To do this I split the information in line_2 on the pipe | again like in step 1 and create two vectors. One contains the state and the other contains the pre and post ranking information for the player. I used gsub to clean and isolate the pre rank score and convert it to an integer. Finally I add the two columns into the dataframe.
pre_rank <- vect_2[seq(2, length(vect_2), by = 10)]
pre_rank <- as.integer(gsub("-.*|.*:|P.*","",pre_rank))
df$pre_rank <- pre_rank
head(df)## player_name state total_points pre_rank
## 1 GARY HUA ON 6.0 1794
## 2 DAKSHESH DARURI MI 6.0 1553
## 3 ADITYA BAJAJ MI 6.0 1384
## 4 PATRICK H SCHILLING MI 5.5 1716
## 5 HANSHI ZUO MI 5.5 1655
## 6 HANSEN SONG OH 5.0 1686Create a column that is an average of the pre-ranks of players that were faced during the tournament. To do this I create a long vector that contains only the player_id of the players faced. NA’s are used in place of players that did not play 7 games and convert the rest to integers. I then create a list of vectors that contain 7 elements for each player. This list is the id’s of the players faced during the tournament. I use this id to retrieve pre_rank value from the df created above. The index values of the dataframe are the same as the player_id. Special consideration is made to use the NA’s as placeholders but remove them when computing the average.
# create the list oponent
all_games <- vect_1[-seq(1, length(vect_1), by = 10)]
all_games <- all_games[-seq(1, length(vect_1), by = 9)]
all_games <- all_games[-seq(1, length(vect_1), by = 8)]
all_games <- gsub("[^0-9]", "", all_games)
all_games <- as.integer(all_games)
# make a list of all oponents for each player
opponents_by_player <- list()
n = 1
for (i in 1:64) {
opponents_by_player[[i]] <- all_games[(n:(n+6))]
n = n + 5
}
# look at player 15-19 (to show NA's)
opponents_by_player[15:19]## [[1]]
## [1] 38 56 6 7 3 34 26
##
## [[2]]
## [1] 34 26 42 33 5 38 NA
##
## [[3]]
## [1] 38 NA 1 3 36 27 7
##
## [[4]]
## [1] 27 7 5 33 3 32 54
##
## [[5]]
## [1] 32 54 44 8 1 27 5# create an empty vector to store average opponenent rank
avg_opponent <- vector()
# loop through list of vecotrs
for (i in (1:(length(opponents_by_player)))){
opponent_rank <- vector()
# nested loop to loop over non null values for each player
for (n in (1:(length(opponents_by_player[[i]][!is.na(opponents_by_player[[i]])])))){
# remove null values from list of opponents
opponent_id_list <- opponents_by_player[[i]][!is.na(opponents_by_player[[i]])]
rank <- df$pre_rank[opponent_id_list[n]]
opponent_rank[n] <- rank
}
avg_opponent[i] <- mean(opponent_rank)
}
# add to dataframe
df$avg_opponent <- as.integer(avg_opponent)
head(df)## player_name state total_points pre_rank avg_opponent
## 1 GARY HUA ON 6.0 1794 1605
## 2 DAKSHESH DARURI MI 6.0 1553 1488
## 3 ADITYA BAJAJ MI 6.0 1384 1545
## 4 PATRICK H SCHILLING MI 5.5 1716 1523
## 5 HANSHI ZUO MI 5.5 1655 1554
## 6 HANSEN SONG OH 5.0 1686 1509