#########################################
# 2014-01-29 Eploratory Data Analysis
#
# Example: Albany Airport daily weather
# reports from December 2013
# KALB_201312.csv 
# (retrieved from 
# http://www.erh.noaa.gov/aly/Climate/Albany/ClimateALB.htm
#########################################

dec<-read.csv("data/KALB_201312.csv",skip=2)
print(names(dec))
# we plot the temperature daily maximum
# on the x-axis the day in the month
# on the y-axis the daily maximum temperature
plot(dec$DY, dec$MAX,main="Albany temperature")
# when plot() has been called we can add
# more into the graph, like a connecting line
lines(dec$DY,dec$MAX,col="red")


res<-readline("hit enter to continue")
# we add the minimum temperatures
lines(dec$DY,dec$MIN,col="blue")

# We can do better but we need to adjust the 
# y-axis range 
# first an empty graph with argument typ='n'

plot(c(1,31),c(-20,70),typ='n',
     main="Albany temperature",
     xlab='day',ylab='temperature [F]')

lines(dec$DY,dec$MAX,col="red")
lines(dec$DY,dec$MIN,col="blue")

res<-readline("hit enter to continue")

print("What was the average December 2013 temperature?")

# first it is common to obtain daily mean 
# temperatures from tmax and tmin as
tday<-(dec$MAX+dec$MIN)/2
lines(dec$DY,tday,col="black")

# the average Dec daily mean temperature is
tmean<-mean(tday)
print(paste("The average December 2013 Temperature is ",tmean))

# NOTE: the computer returns long decimals in the result
# We will learn soon that it is reasonable to
# round the result to one or two digits.
tmean<-round(tmean,1)

print(paste("The average December 2013 Temperature is ",tmean))

lines(c(1,31),c(tmean,tmean),col="black",lwd=2)


dev.copy2pdf(file="figures/KALB_temp_201312.pdf")

res<-readline("hit enter to continue")


###################################################
# now back to our probability and events
# How many freezing days were in December?
# We use daily max temperatures
###################################################

# R allows for a number of operations to compare
# data values (x==1 x<1 X>1) 
# the result is TRUE or FALSE which is internally
# also for R a number of 1 or 0
# I use this exercise to introduce the two most
# important concepts in writing computer programs
# loops and if-else statements
# a new set of brackets is used here to enclose
# a block of commands that are to be executed in 
# loops, if-else-statements
# NOTE:
# text lines wihtin the loop are (and should always be) 
# indented.
# So is then the if block, which makes the code
# easier to read and understand.

count<-0
for (n in 1:31) {
  if (dec$MAX[n]<=32.0) {
    count<-count+1
    print (paste(count,": Dec. ",n," was below 32F"))
  } else {
    # empty block, nothing to do in this case
  }
}

print(paste("Frequentist's approach to probability 
             estimation of freezing events"))
print(paste("sample size n=",n, "freezing events: ",count))
P1<-round(count/n,2)
P2<-1-P1
print(paste("estimated probability for freezing events :",P1))
print(paste("The probability of the complement event:",P2))
       
            
      

###################################################
# INFO:
# the function read.csv()
# is good for reading data tables where 
# columns are delimited with a colons
# EXCEL spreadsheets can be exported to csv
# format.
# the argument in the function call "skip=2"
# tells the function to ignore the 
# first two lines while reading the data table
# print the names of the data columns
# The third line in KALB_201312.csv contains
# useful names for the data columns
# which we can use in R
# (please take a look at the file KALB_201312.csv
# how it looks)