#
# Written by:
# -- 
# John L. Weatherwax                2009-04-21
# 
# email: wax@alum.mit.edu
# 
# Please send comments and especially bug reports to the
# above email address.
#
#-----

source('wilcoxon_rank_sum_interval.R')
source('wilcoxon_rank_sum_test.R')
source('friedmans_test.R')

# Ex 15.28: EPage 622; Anwsers EPage 708 
#
DF = read.csv( "../../Data/CH15/ex15-28.txt", header=TRUE, quote="'" )
x_data = DF[ DF$'diet' == 'Potato', ]$cholesterol
y_data = DF[ DF$'diet' == 'Rice', ]$cholesterol
out = wilcoxon_rank_sum_test( x_data, y_data ) 
print( out$w )

# Ex 15.29: EPage 621; Anwsers EPage 708 
#
DF = read.csv( "../../Data/CH15/ex15-29.txt", header=TRUE, quote="'" )
out = kruskal_wallis_test( DF, 'Year', 'cancellations', alpha=0.05 ) # using Kruskal-Wallis
print( out )

out = friedmans_test( DF, 'Salesperson', 'Year', 'cancellations', alpha=0.05 ) # using Friedmans' test
print( out )

# Ex 15.30: EPage 621; Anwsers EPage 708 
#
DF = read.csv( "../../Data/CH15/ex15-30.txt", header=TRUE, quote="'" )
out = kruskal_wallis_test( DF, 'treatments', 'phosphorus', alpha=0.05 ) # using Kruskal-Wallis
print( out )

# Ex 15.31: EPage 618; Anwsers EPage 708; Exercise 11 EPage 613
#
DF = read.csv( "../../Data/CH15/ex15-30.txt", header=TRUE, quote="'" )
x_data = DF[ DF$treatments == 'II', ]$phosphorus
y_data = DF[ DF$treatments == 'III', ]$phosphorus
c = 22
out = wilcoxon_rank_sum_interval( x_data, y_data, c )
print( out )

# Ex 15.32: EPage 622; Anwsers EPage 708 
#
DF = read.csv( "../../Data/CH15/ex15-32.txt", header=TRUE, quote="'" )
x_data = DF[ DF$'gait.type' == 'Diagonal', ]$times
y_data = DF[ DF$'gait.type' == 'Lateral', ]$times
out = wilcoxon_rank_sum_test( x_data, y_data ) 
print( out$w )

# Part (b) 95% CI: 
ci_outer = outer( x_data, y_data, FUN="-" )
ci_sorted = sort( as.vector( ci_outer ) )
ci_sorted[8]
ci_sorted[35] 

# Ex 15.33/34: EPage 622; Anwsers EPage 708 
#
DF = read.csv( "../../Data/CH15/ex15-33.txt", header=TRUE, quote="'" )
num_gt_25 = sum( DF > 25 ) 

tilde_mus = unique( DF$Observations )
thresholds = tilde_mus
did_not_reject = c()
for( t in thresholds ){
  num_gt = sum( DF > t ) 
  if( (num_gt > 5) & (num_gt < 15) ){
    did_not_reject = c( did_not_reject, t )
  }
} 
print( did_not_reject )