Code Base

Approach

For this project, I will use the chess tournament text file from the course. It is not arranged like a normal CSV because each player uses two lines. The first line has the player’s number, name, points, and round results. The second line has the state and pre-rating.

I will read the file into R and find the lines that start with a player number. I will split those lines at the vertical bars and collect the name, points, and opponent numbers. I will use the next line for the state and pre-rating.

Next, I will match each opponent number with that player’s pre-rating and calculate the average. I will leave out byes and unplayed rounds because they do not have an opponent.

The final dataframe will contain the player’s name, state, total points, pre-rating, and average opponent pre-rating. I will check that there are 64 players, look for missing or duplicate values, compare Gary Hua’s result with the example, and export the table to CSV.

Setup

# tidyverse: data wrangling, parsing, and charts
library(tidyverse)
# stringr: text matching and cleanup
library(stringr)
# knitr: clean HTML table output
library(knitr)

Load the Tournament File

# Look for the tournament file in common project locations first.
# This avoids using a path that only works on one computer.
local_files <- c(
  "tournamentinfo.txt",
  file.path("data", "tournamentinfo.txt"),
  file.path("Project 1", "tournamentinfo.txt")
)

input_file <- local_files[file.exists(local_files)][1]

# Online fallback: public copy of the same DATA 607 tournament dataset.
data_url <- paste0(
  "https://raw.githubusercontent.com/longflin/DATA-607/",
  "refs/heads/main/Project%201/tournamentinfo.txt"
)

if (!is.na(input_file)) {
  raw_lines <- readLines(input_file, warn = FALSE)
  data_source <- input_file
} else {
  raw_lines <- tryCatch(
    readLines(data_url, warn = FALSE),
    error = function(e) {
      stop(
        "Tournament data could not be loaded. ",
        "Put tournamentinfo.txt in the same folder as this QMD ",
        "or in a data/ folder, then render again. ",
        "The online fallback also could not be reached."
      )
    }
  )
  data_source <- data_url
}

cat("Data source:", data_source)

Data source: https://raw.githubusercontent.com/longflin/DATA-607/refs/heads/main/Project%201/tournamentinfo.txt

The code checks for a local copy first. If the file is not there, it tries an online copy of the same tournament dataset. This fixes the original working-directory problem while still allowing the course file to be used when it is available.

Find the Player Records

# Find lines that begin with a player number followed by a vertical bar
player_line_index <- which(
  str_detect(raw_lines, "^\\s*\\d+\\s*\\|")
)

length(player_line_index)

[1] 64

The first line for each player starts with a player number. I use that pattern to find the player records instead of depending on fixed row numbers.

Parse the Player Information

# Remove extra spaces around parsed text fields
clean_field <- function(x) {
  str_squish(x)
}

# Pull a 3- or 4-digit chess rating from the rating field
parse_rating <- function(x) {
  rating_match <- str_match(x, "(\\d{3,4})")[, 2]
  as.numeric(rating_match)
}

# Parse every player record and combine the results into one dataframe
players <- map_dfr(player_line_index, function(i) {

  line1 <- str_split(raw_lines[i], "\\|", simplify = TRUE)
  line2 <- str_split(raw_lines[i + 1], "\\|", simplify = TRUE)

  player_no <- as.integer(clean_field(line1[1]))
  player_name <- clean_field(line1[2])
  total_points <- parse_number(clean_field(line1[3]))

  state <- clean_field(line2[1])
  pre_rating <- parse_rating(clean_field(line2[2]))

  round_fields <- if (ncol(line1) >= 4) line1[1, 4:ncol(line1)] else character(0)

  opponent_numbers <- str_match(round_fields, "(\\d+)")[, 2] |>
    as.integer()

  tibble(
    player_no = player_no,
    name = player_name,
    state = state,
    total_points = total_points,
    pre_rating = pre_rating,
    opponent_numbers = list(opponent_numbers)
  )
})

This separates the fields from the two-line record and stores each player’s opponent numbers for the next step.

Check the Parsed Data

dim(players)

[1] 64  6

head(players)

# A tibble: 6 × 6
  player_no name                state total_points pre_rating opponent_numbers
      <int> <chr>               <chr>        <dbl>      <dbl> <list>          
1         1 GARY HUA            ON             6         1544 <int [8]>       
2         2 DAKSHESH DARURI     MI             6         1459 <int [8]>       
3         3 ADITYA BAJAJ        MI             6         1495 <int [8]>       
4         4 PATRICK H SCHILLING MI             5.5       1261 <int [8]>       
5         5 HANSHI ZUO          MI             5.5       1460 <int [8]>       
6         6 HANSEN SONG         OH             5         1505 <int [8]>       

sum(is.na(players$name))

[1] 0

sum(is.na(players$pre_rating))

[1] 0

sum(duplicated(players$player_no))

[1] 0

The project description I am using expects 64 players. I also check for missing names, missing ratings, and duplicate player numbers before calculating opponent averages.

stopifnot(nrow(players) == 64)
stopifnot(!anyDuplicated(players$player_no))

If one of these checks fails, the code stops instead of silently producing a bad result.

Calculate Average Opponent Pre-Rating

# Create a lookup table: player number -> pre-rating
rating_lookup <- players |>
  select(player_no, opponent_pre_rating = pre_rating)

# Expand opponent numbers, join opponent ratings, then average by player
final_data <- players |>
  select(
    player_no,
    name,
    state,
    total_points,
    pre_rating,
    opponent_numbers
  ) |>
  unnest_longer(
    opponent_numbers,
    values_to = "opponent_no",
    keep_empty = TRUE
  ) |>
  filter(!is.na(opponent_no)) |>
  left_join(
    rating_lookup,
    by = c("opponent_no" = "player_no")
  ) |>
  group_by(
    player_no,
    name,
    state,
    total_points,
    pre_rating
  ) |>
  summarise(
    average_opponent_pre_rating =
      round(mean(opponent_pre_rating, na.rm = TRUE)),
    .groups = "drop"
  ) |>
  arrange(player_no)

Byes and unplayed rounds do not produce an opponent number, so they are left out of the average.

Final Table

# Keep and rename only the five fields used in the final result
project1_table <- final_data |>
  transmute(
    Name = name,
    State = state,
    `Total Number of Points` = total_points,
    `Pre-Rating` = pre_rating,
    `Average Pre Tournament Chess Rating of Opponents` =
      average_opponent_pre_rating
  )

kable(
  project1_table,
  caption = "Project 1 Chess Tournament Results"
)

Project 1 Chess Tournament Results

Name	State	Total Number of Points	Pre-Rating	Average Pre Tournament Chess Rating of Opponents
GARY HUA	ON	6.0	1544	1233
DAKSHESH DARURI	MI	6.0	1459	1266
ADITYA BAJAJ	MI	6.0	1495	1417
PATRICK H SCHILLING	MI	5.5	1261	1493
HANSHI ZUO	MI	5.5	1460	1309
HANSEN SONG	OH	5.0	1505	1452
GARY DEE SWATHELL	MI	5.0	1114	1475
EZEKIEL HOUGHTON	MI	5.0	1514	1377
STEFANO LEE	ON	5.0	1495	1323
ANVIT RAO	MI	5.0	1415	1351
CAMERON WILLIAM MC LEMAN	MI	4.5	1258	1444
KENNETH J TACK	MI	4.5	1268	1487
TORRANCE HENRY JR	MI	4.5	1508	1410
BRADLEY SHAW	MI	4.5	1013	1470
ZACHARY JAMES HOUGHTON	MI	4.5	1561	1325
MIKE NIKITIN	MI	4.0	1029	1441
RONALD GRZEGORCZYK	MI	4.0	1029	1455
DAVID SUNDEEN	MI	4.0	1134	1444
DIPANKAR ROY	MI	4.0	1486	1409
JASON ZHENG	MI	4.0	1452	1466
DINH DANG BUI	ON	4.0	1549	1427
EUGENE L MCCLURE	MI	4.0	1240	1426
ALAN BUI	ON	4.0	1503	1388
MICHAEL R ALDRICH	MI	4.0	1346	1383
LOREN SCHWIEBERT	MI	3.5	1248	1395
MAX ZHU	ON	3.5	1513	1313
GAURAV GIDWANI	MI	3.5	1447	1440
SOFIA ADINA STANESCU-BELLU	MI	3.5	1488	1397
CHIEDOZIE OKORIE	MI	3.5	1532	1487
GEORGE AVERY JONES	ON	3.5	1257	1523
RISHI SHETTY	MI	3.5	1513	1382
JOSHUA PHILIP MATHEWS	ON	3.5	1407	1471
JADE GE	MI	3.5	1469	1467
MICHAEL JEFFERY THOMAS	MI	3.5	1505	1434
JOSHUA DAVID LEE	MI	3.5	1460	1496
SIDDHARTH JHA	MI	3.5	1477	1421
AMIYATOSH PWNANANDAM	MI	3.5	1548	1498
BRIAN LIU	MI	3.0	1510	1369
JOEL R HENDON	MI	3.0	1292	1396
FOREST ZHANG	MI	3.0	1489	1400
KYLE WILLIAM MURPHY	MI	3.0	1576	1309
JARED GE	MI	3.0	1446	1465
ROBERT GLEN VASEY	MI	3.0	1410	1468
JUSTIN D SCHILLING	MI	3.0	1532	1266
DEREK YAN	MI	3.0	1537	1489
JACOB ALEXANDER LAVALLEY	MI	3.0	1549	1416
ERIC WRIGHT	MI	2.5	1253	1413
DANIEL KHAIN	MI	2.5	1436	1403
MICHAEL J MARTIN	MI	2.5	1253	1488
SHIVAM JHA	MI	2.5	1477	1461
TEJAS AYYAGARI	MI	2.5	1520	1443
ETHAN GUO	MI	2.5	1491	1417
JOSE C YBARRA	MI	2.0	1257	1430
LARRY HODGE	MI	2.0	1283	1371
ALEX KONG	MI	2.0	1541	1442
MARISA RICCI	MI	2.0	1467	1438
MICHAEL LU	MI	2.0	1511	1379
VIRAJ MOHILE	MI	2.0	1470	1474
SEAN M MC CORMICK	MI	2.0	1284	1464
JULIA SHEN	MI	1.5	1457	1461
JEZZEL FARKAS	ON	1.5	1577	1384
ASHWIN BALAJI	MI	1.0	1521	1541
THOMAS JOSEPH HOSMER	MI	1.0	1505	1419
BEN LI	MI	1.0	1500	1363

This is the table I will export for the project.

Quality Checks

nrow(project1_table)

[1] 64

colSums(is.na(project1_table))

                                            Name 
                                               0 
                                           State 
                                               0 
                          Total Number of Points 
                                               0 
                                      Pre-Rating 
                                               0 
Average Pre Tournament Chess Rating of Opponents 
                                               0 

sum(duplicated(project1_table$Name))

[1] 0

summary(project1_table)

        Name          State    Total Number of Points   Pre-Rating  
 Length   :64   Length   :64   Min.   :1.000          Min.   :1013  
 N.unique :64   N.unique : 3   1st Qu.:2.500          1st Qu.:1290  
 N.blank  : 0   N.blank  : 0   Median :3.500          Median :1474  
 Min.nchar: 6   Min.nchar: 2   Mean   :3.438          Mean   :1416  
 Max.nchar:26   Max.nchar: 2   3rd Qu.:4.000          3rd Qu.:1512  
                               Max.   :6.000          Max.   :1577  
 Average Pre Tournament Chess Rating of Opponents
 Min.   :1233                                    
 1st Qu.:1384                                    
 Median :1426                                    
 Mean   :1418                                    
 3rd Qu.:1465                                    
 Max.   :1541                                    

I use these checks to make sure the final table has the expected number of rows and to catch missing or duplicate values.

Gary Hua Check

gary_hua_check <- project1_table |>
  filter(str_to_lower(Name) == "gary hua")

gary_hua_check

# A tibble: 1 × 5
  Name     State `Total Number of Points` `Pre-Rating` Average Pre Tournament …¹
  <chr>    <chr>                    <dbl>        <dbl>                     <dbl>
1 GARY HUA ON                           6         1544                      1233
# ℹ abbreviated name: ¹​`Average Pre Tournament Chess Rating of Opponents`

I included this check because the approach says I will compare Gary Hua’s result with the course example. I will make that comparison from the rendered output rather than typing a result into the document before the code runs.

Exploratory Charts

These charts are secondary to the required table. I kept them simple so they help me inspect the cleaned data without taking over the project.

Pre-Rating Distribution

ggplot(project1_table, aes(x = `Pre-Rating`)) +
  geom_histogram(
    bins = 12,
    fill = "steelblue",
    color = "white"
  ) +
  labs(
    title = "Distribution of Player Pre-Ratings",
    x = "Pre-Rating",
    y = "Number of Players"
  ) +
  theme_minimal()

The histogram gives a quick view of how the player ratings are distributed.

Points by Player

# Order players by points and use pre-rating to break ties.
# The taller figure gives all player names enough room to be readable.
points_plot <- project1_table |>
  arrange(`Total Number of Points`, `Pre-Rating`) |>
  mutate(Name = factor(Name, levels = Name))

ggplot(
  points_plot,
  aes(x = `Total Number of Points`, y = Name)
) +
  geom_segment(
    aes(
      x = 0,
      xend = `Total Number of Points`,
      y = Name,
      yend = Name
    ),
    color = "grey75",
    linewidth = 0.6
  ) +
  geom_point(
    size = 3,
    color = "steelblue"
  ) +
  labs(
    title = "Tournament Points by Player",
    subtitle = "Players are ordered by total tournament points",
    x = "Total Points",
    y = NULL
  ) +
  scale_x_continuous(
    breaks = seq(
      0,
      max(points_plot$`Total Number of Points`, na.rm = TRUE),
      by = 1
    )
  ) +
  theme_minimal(base_size = 11) +
  theme(
    axis.text.y = element_text(size = 8),
    panel.grid.major.y = element_blank(),
    panel.grid.minor = element_blank(),
    plot.title = element_text(face = "bold")
  )

The taller lollipop chart makes the player names readable. Players are ordered by total points, with pre-rating used to break ties. This chart is supporting analysis; the final project table remains the main output. ### Player Rating and Opponent Rating

ggplot(
  project1_table,
  aes(
    x = `Pre-Rating`,
    y = `Average Pre Tournament Chess Rating of Opponents`
  )
) +
  geom_point(
    alpha = 0.75,
    color = "steelblue"
  ) +
  labs(
    title = "Player Rating vs. Average Opponent Rating",
    x = "Player Pre-Rating",
    y = "Average Opponent Pre-Rating"
  ) +
  theme_minimal()

This scatterplot compares two numeric variables from the final dataset.

Export the CSV

# Save the final table as a CSV file
write_csv(
  project1_table,
  "project1_chess_results.csv"
)

The exported CSV contains only the five requested result columns.

AI Use

I used GitHub Copilot in Visual Studio Code to help review parts of the R code, troubleshoot errors, and improve code comments. I reviewed the suggestions, kept the code I understood, and checked the results when running the project. The analysis and final results come from the R code and tournament data used in this project.

References

DATA 607 chess tournament dataset (public copy used only as a fallback):
https://raw.githubusercontent.com/longflin/DATA-607/refs/heads/main/Project%201/tournamentinfo.txt

Wickham, H., Cetinkaya-Rundel, M., & Grolemund, G. R for Data Science (2e).

R Graph Gallery. Lollipop plot examples.

R Graph Gallery. Histogram examples.