#
# Read in pronoun frequencies by age and sex, from PCC dataset:
download.file("http://ling.upenn.edu/courses/ling005/FP.txt", "FP.txt")
Pronouns = read.table("FP.txt")
#################################################################
# Pronouns is a data.frame with 90 rows (45 ages for males, 45 ages for females)
#  and 9 columns ("AGE" "SEX" "FPS" "SP"  "FPP" "TPF" "TPM" "TPN" "TPP") where
#  FPS is First Person Singular ("I", "me", "my", "mine")
#  SP is Second Person ("you", "your", "yours")
#  FPP is First Person Plural ("we", "us", "our", "ours")
#  TPF is Third Person Feminine ("she", "her", "hers")
#  TPM is Third Person Masculine ("he", "him", "his")
#  TPN is Third Person Neuter ("it", "its")
#  TPP is Third Person Plural ("they", "them", "their", "theirs")
#
# "AGE" is 15-41 and 43-60
#    leaving out 42 because it was the default assigns to subjects of unknown age
################################################################
# a useful boolean:
IsMale = Pronouns[,"SEX"]==0
IsFemale = Pronouns[,"SEX"]==1
# The set of ages:
Ages = Pronouns[IsFemale,"AGE"]
#####################################
# Tell R to ask us before making a new plot
par(ask=TRUE)
######################################
# Let's plot FPS pronouns as a function of age,
#         averaging males and females:
allFPS = (Pronouns[IsMale,"FPS"]+Pronouns[IsFemale,"FPS"])/2
plot(Ages, allFPS, type="b", col="magenta", pch="I",
     xlab="Age", ylab="Frequency (per million words)",
     main="First Person Singular Pronouns (PPC Facebook dataset)")
#
# Now FPP:
allFPP = (Pronouns[IsMale,"FPP"]+Pronouns[IsFemale,"FPP"])/2
plot(Ages, allFPP, type="b", col="forestgreen", pch="W",
     xlab="Age", ylab="Frequency (per million words)",
     main="First Person Plural Pronouns (PPC Facebook dataset)")
# Now both on the same plot:
yrange = range(c(allFPS,allFPP))
plot(Ages, allFPS, type="b", col="magenta", pch="I",
     xlab="Age", ylab="Frequency (per million words)", ylim=yrange,
     main="First Person Pronouns (PPC Facebook dataset)")
points(Ages, allFPP, type="b", col="forestgreen", pch="W")
# Now the ratio:
plot(Ages, allFPS/allFPP, type="b", pch="E", col="red",
     xlab="Ages", ylab="Ratio",
     main="First person singular / First person plural pronouns\n\"me/we ratio\"")
###############################################################
# Let's pull out the frequency of masculine vs. feminine pronoun references
#         for male vs. female Facebookers of different ages:
M2M = Pronouns[IsMale,"TPM"]    # males referring to males
M2F = Pronouns[IsMale,"TPF"]      # males referring to females
F2M = Pronouns[IsFemale,"TPM"] # females referring to males
F2F = Pronouns[IsFemale,"TPF"]  # females referring to females
#
# Now we plot the rate of same-sex vs. cross-sex reference
#     as a function of age:
SameSexRef = (M2M+F2F)/2
CrossSexRef = (M2F+F2M)/2
yrange = range(c(SameSexRef,CrossSexRef))
plot(Ages, SameSexRef, type="b", pch="=", col="blue",
     xlab="Age", ylab="Frequency (per million words)", ylim=yrange,
     main="Same Sex reference (blue =) and Cross Sex reference (red x)")
points(Ages, CrossSexRef, type="b", pch="x", col="red")