# importing data
sp <- read.csv("/Users/deborah/Desktop/hhelianthoides.csv")
	
# the data
#sp$headId = identification number assigned to flower head
#sp$row = which row of styles (1=bottom row)
#sp$treatment = pollination treatment applied
#sp$styleDate = date when styles emerged
#sp$dateTreat = date when treatment was first applied
#sp$styleShriv = whether styles were shriveled on first visit after treatment (0=no, 1-yes)
#sp$dateShriv = date when styles were observed to have shriveled
#sp$daysPerstLong = large estimate of how long styles persisted
#sp$daysPerstAvg = middle estimate of how long styles persisted
#sp$daysPerstShort = small estimate of how long styles persisted
#sp$lastDayFlw = last day flower head put out anthers
#sp$endShrivLong = large estimate of how many days styles persisted after last day of flowering (0=shriveled on last day)
#sp$endShrivAvg = middle estimate of how many days styles persisted after last day of flowering
#sp$endShrivShort = small estimate of how many days styles persisted after last day of flowering
#sp$shrivNotes = notes about shriveling
#sp$repollinated = date when pollination treatment was reapplied
#sp$notes = other notes

# checking the data file
str(sp) #structure
sp$headId
levels(sp$treatment)

# how many rows received each treatment
table(sp$treatment)

# how many rows do heads have
hist(sp$row, breaks= 0:20) #histogram of "how many heads have a row with this number", not of total number of rows on each head

# phenology
#This is not a representative sample of the population. For my experiment, I selected heads that flowered later than most.
# how do I make a bar graph?
table(sp$styleDate) #how many rows of styles came out each day
fs <- subset(sp, row == 1)
table(fs$styleDate) #how many heads started flowering one day before
table(fs$lastDayFlw) #how many heads finished flowering each day

# subsets for each treatment
sc <- subset(sp, treatment == "control")
so <- subset(sp, treatment == "outcross")
ss <- subset(sp, treatment == "self")

# style persistence after treatment
summary(sp$styleShriv)
summary(sc$styleShriv)
summary(so$styleShriv)
summary(ss$styleShriv)
#How do I compare three proportions in R?

# length of style persistence
summary(sp$daysPerstAvg) #overall
hist(sp$daysPerstAvg, breaks= 0:18) 
summary(sc$daysPerstAvg) #control
quartz()
hist(sc$daysPerstAvg, breaks= 0:18)
summary(so$daysPerstAvg) #outcross
quartz()
hist(so$daysPerstAvg, breaks= 0:18)
summary(ss$daysPerstAvg) #self
quartz()
hist(ss$daysPerstAvg, breaks= 0:18)
summary(aov(sp$daysPerstAvg~sp$treatment))
#Can I add lines to show the high and low estimates, or should I just graph these separately?

# length of style persistence after last day of flowering
summary(sp$endShrivAvg) #overall
quartz()
hist(sp$endShrivAvg, breaks= -12:12) 
summary(sc$endShrivAvg) #control
quartz()
hist(sc$endShrivAvg, breaks= -12:12)
summary(so$endShrivAvg) #outcross
quartz()
hist(so$endShrivAvg, breaks= -12:12)
summary(ss$endShrivAvg) #self
quartz()
hist(ss$endShrivAvg, breaks= -12:12)
summary(aov(sp$endShrivAvg~sp$treatment))
#Can I add lines to show the high and low estimates, or should I just graph these separately?



# re-analysis excluding 23-Jul-11

# subsets excluding 23-Jul-11
rp <- subset(sp, dateTreat != "21-Jul-11")
rc <- subset(sc, dateTreat != "21-Jul-11")
ro <- subset(so, dateTreat != "21-Jul-11")
rs <- subset(ss, dateTreat != "21-Jul-11")

# style persistence after treatment
summary(rp$styleShriv)
summary(rc$styleShriv)
summary(ro$styleShriv)
summary(rs$styleShriv)
#How do I compare three proportions in R?

# length of style persistence
summary(rp$daysPerstAvg) #overall
quartz()
hist(rp$daysPerstAvg, breaks= 0:15) 
summary(rc$daysPerstAvg) #control
quartz()
hist(rc$daysPerstAvg, breaks= 0:15)
summary(ro$daysPerstAvg) #outcross
quartz()
hist(ro$daysPerstAvg, breaks= 0:15)
summary(rs$daysPerstAvg) #self
quartz()
hist(rs$daysPerstAvg, breaks= 0:15)
summary(aov(rp$daysPerstAvg~rp$treatment))
#Can I add lines to show the high and low estimates, or should I just graph these separately?

# length of style persistence after last day of flowering
summary(rp$endShrivAvg) #overall
quartz()
hist(rp$endShrivAvg, breaks= -12:12) 
summary(rc$endShrivAvg) #control
quartz()
hist(rc$endShrivAvg, breaks= -12:12)
summary(ro$endShrivAvg) #outcross
quartz()
hist(ro$endShrivAvg, breaks= -12:12)
summary(rs$endShrivAvg) #self
quartz()
hist(rs$endShrivAvg, breaks= -12:12)
summary(aov(rp$endShrivAvg~rp$treatment))
#Can I add lines to show the high and low estimates, or should I just graph these separately?



# analysis of first four rows only

# subsets of first four rows only
ep <- subset(sp, row < 5)
ec <- subset(sc, row < 5)
eo <- subset(so, row < 5)
es <- subset(ss, row < 5)

# style persistence after treatment
summary(ep$styleShriv)
summary(ec$styleShriv)
summary(eo$styleShriv)
summary(es$styleShriv)
#How do I compare three proportions in R?

# length of style persistence
summary(ep$daysPerstAvg) #overall
quartz()
hist(ep$daysPerstAvg, breaks= 0:15) 
summary(ec$daysPerstAvg) #control
quartz()
hist(ec$daysPerstAvg, breaks= 0:15)
summary(eo$daysPerstAvg) #outcross
quartz()
hist(eo$daysPerstAvg, breaks= 0:15)
summary(es$daysPerstAvg) #self
quartz()
hist(es$daysPerstAvg, breaks= 0:15)
summary(aov(ep$daysPerstAvg~ep$treatment))
#Can I add lines to show the high and low estimates, or should I just graph these separately?

# length of style persistence after last day of flowering
summary(ep$endShrivAvg) #overall
quartz()
hist(ep$endShrivAvg, breaks= -12:5) 
summary(ec$endShrivAvg) #control
quartz()
hist(ec$endShrivAvg, breaks= -12:5)
summary(eo$endShrivAvg) #outcross
quartz()
hist(eo$endShrivAvg, breaks= -12:5)
summary(es$endShrivAvg) #self
quartz()
hist(es$endShrivAvg, breaks= -12:5)
summary(aov(ep$endShrivAvg~ep$treatment))
#Can I add lines to show the high and low estimates, or should I just graph these separately?