R Notebook

library(tidyverse)

d18th <- read_csv("18th-data.csv") %>% filter(`Ok for festschrift?`!="yess")

── Column specification ──────────────────────────────────────────────────────────────────────────────────────────────────
cols(
  .default = col_character(),
  `Earliest letter` = col_double(),
  Frequency = col_double(),
  `Total DF after` = col_double(),
  `Total DF before` = col_double(),
  word_daterange_start = col_double(),
  sense_daterange_start = col_double(),
  MultiSenders = col_logical(),
  MultiRec = col_logical(),
  Year = col_double(),
  YearUncertain = col_logical(),
  WordCount = col_double(),
  AddressFormula = col_logical(),
  ClosingFormula = col_logical(),
  CorrespondentNotes = col_logical(),
  LetterNumber = col_logical(),
  Copyright = col_logical(),
  Complete = col_logical(),
  Updated = col_datetime(format = ""),
  `non-normalized` = col_double(),
  words = col_double()
  # ... with 16 more columns
)
ℹ Use `spec()` for the full column specifications.

s18th <- read_csv("18th-sampleinfo.csv")

── Column specification ──────────────────────────────────────────────────────────────────────────────────────────────────
cols(
  .default = col_character(),
  MultiSenders = col_logical(),
  MultiRec = col_logical(),
  Year = col_double(),
  YearUncertain = col_logical(),
  WordCount = col_double(),
  AddressFormula = col_logical(),
  ClosingFormula = col_logical(),
  LetterNumber = col_logical(),
  Copyright = col_logical(),
  Complete = col_logical(),
  Updated = col_datetime(format = ""),
  `non-normalized` = col_double(),
  words = col_double(),
  SenderYBirth = col_double(),
  SenderYDeath = col_double(),
  SenderReligion = col_logical(),
  SenderSentLettcont = col_logical(),
  SenderRecLettcont = col_logical(),
  SenderUpdated = col_datetime(format = ""),
  RecipientYBirth = col_double()
  # ... with 9 more columns
)
ℹ Use `spec()` for the full column specifications.

d17th <- read_csv("17th-data.csv") %>% filter(`Ok for festschrift?`!="yess")

── Column specification ──────────────────────────────────────────────────────────────────────────────────────────────────
cols(
  .default = col_character(),
  `Earliest letter` = col_double(),
  Frequency = col_double(),
  `Total DF after` = col_double(),
  `Total DF before` = col_double(),
  word_daterange_start = col_double(),
  sense_daterange_start = col_double(),
  MultiSenders = col_logical(),
  MultiRec = col_logical(),
  Year = col_double(),
  YearUncertain = col_logical(),
  WordCount = col_double(),
  AddressFormula = col_logical(),
  ClosingFormula = col_logical(),
  CorrespondentNotes = col_logical(),
  LetterNumber = col_logical(),
  Copyright = col_logical(),
  Complete = col_logical(),
  Updated = col_datetime(format = ""),
  `non-normalized` = col_double(),
  words = col_double()
  # ... with 10 more columns
)
ℹ Use `spec()` for the full column specifications.

s17th <- read_csv("17th-sampleinfo.csv")

── Column specification ──────────────────────────────────────────────────────────────────────────────────────────────────
cols(
  .default = col_character(),
  MultiSenders = col_logical(),
  MultiRec = col_logical(),
  Year = col_double(),
  YearUncertain = col_logical(),
  WordCount = col_double(),
  AddressFormula = col_logical(),
  ClosingFormula = col_logical(),
  LetterNumber = col_logical(),
  Copyright = col_logical(),
  Complete = col_logical(),
  Updated = col_datetime(format = ""),
  `non-normalized` = col_double(),
  words = col_double(),
  SenderYBirth = col_double(),
  SenderYDeath = col_double(),
  SenderUpdated = col_datetime(format = ""),
  RecipientYBirth = col_double(),
  RecipientYDeath = col_double(),
  RecipientUpdated = col_datetime(format = ""),
  SenderAge = col_double()
  # ... with 3 more columns
)
ℹ Use `spec()` for the full column specifications.

g <- function(d1,d2,grouping) {
  grouping = enquo(grouping)
  return(d1 %>% group_by(!!grouping) %>% summarize(tokens=n(),lemmas=n_distinct(Lemma),authors=n_distinct(Sender),letters=n_distinct(Letters)) %>% right_join(d2 %>% group_by(!!grouping) %>% summarize(nauthors=n_distinct(Sender),nwords=sum(WordCount),nletters=n()),by=c(quo_name(grouping))) %>% mutate(tokens_per_10k_words=tokens/nwords*10000,letter_proportion=letters/nletters,author_proportion=authors/nauthors,lemmas_per_10k_words=lemmas/nwords*10000) %>% arrange(desc(tokens_per_10k_words)) %>% relocate(!!grouping,tokens_per_10k_words,lemmas_per_10k_words,letter_proportion,author_proportion,lemmas,letters,authors,nwords,nletters,nauthors))
}

Distinct values in each column

17th

d17th %>% summarize_all(~length(unique(.)))

18th

d18th %>% summarize_all(~length(unique(.)))

Sex

17th

g(d17th,s17th,SenderSex)

18th

g(d18th,s18th,SenderSex)

Rank

17th

g(d17th %>% mutate(SenderRank=str_sub(SenderRank,1,1)),s17th %>% mutate(SenderRank=str_sub(SenderRank,1,1)),SenderRank)

18th

g(d18th %>% mutate(SenderRank=str_sub(SenderRank,1,1)),s18th %>% mutate(SenderRank=str_sub(SenderRank,1,1)),SenderRank)

Relationship

17th

g(d17th,s17th,RelCode)

18th

g(d18th,s18th,RelCode)

Education

17th

g(d17th %>% mutate(SenderEduCode=str_sub(SenderEduCode,1,1)),s17th %>% mutate(SenderEduCode=str_sub(SenderEduCode,1,1)),SenderEduCode)

18th

g(d18th %>% mutate(SenderEduCode=str_sub(SenderEduCode,1,1)),s18th %>% mutate(SenderEduCode=str_sub(SenderEduCode,1,1)),SenderEduCode)

Region

17th

g(d17th,s17th,SenderRegion)

18th

g(d18th,s18th,SenderRegion)

Sender

17th

g(d17th,s17th,Sender)

18th

g(d18th,s18th,Sender)