Homework Problem 3 Solution                                       9/13/2017

## (a) 

> framfix = function(infram) {
 # locate the index of any character columns
      charcols = which( sapply(infram, is.character) )
 # Recode NA to "Mis" for missing values in character columns
      for (i in charcols) infram[ infram[,i]=="" ,i] = "Mis"  
   
 # locate the columns named "age" or "Age" 
	 agecol = which( names(infram) %in% c("age","Age") )
 # use ceiling function to determine k, recode age based on rule
      if(length(agecol)>0) 
            for (i in agecol) infram[,i] = ifelse(infram[,i]>0, 
	             5*ceiling(infram[,i]/5)-2.5, 0)
      
 # use "which" function to locate columns with >30% missing values
      manymiss = which( sapply(infram, function(col) 
            mean(is.na(col)) >	0.3 ) )
 # if there are any such columns, delete them.
      if(length(manymiss>0)) newfram = 
            infram[, -manymiss] else newfram=infram
      
 # determine the index of any numeric column
      numcol = which( sapply(newfram, is.numeric) )
 # apply "quantile" with argument "probs" to each numeric column
      Quant = apply( data.matrix(newfram[,numcol]), 2, quantile, 
	             probs=c(.1,.25,.5,.75,.9), na.rm=T)
      
 # determine the index of any logical column
      logcol = which( sapply(newfram, is.logical) )
 # convert boolean values to numeric values in each logical column
      for (i in logcol) newfram[,i] = as.numeric(newfram[,i])
      
      numcol = which( sapply(newfram, is.numeric) )
      tmp = data.matrix(newfram[,numcol])
      list( newfram = newfram, Quant=Quant, 
 # indices of rows containing missing values
            Rows = as.numeric(which(apply(tmp,1, 
                function(row) sum(is.na(row))>0)) ),
 # names of columns containing missing values
	       Cols = names(newfram)[numcol][ apply(tmp,2,
                function(col) sum(is.na(col))>0) ] )
	}			
		
		
## (b)
 
# (b) Part 1
> tmp = framfix(nepali)
> names(tmp)
[1] "newfram" "Quant"   "Rows"    "Cols"  
> tmp[[3]]
  [1]   5  10  19  25  29  82  84  93 151 167 168 169 170 172 173 174 175 182
 [19] 183 193 199 280 304 305 309 310 331 429 462 463 464 465 474 475 484 497
 [37] 498 499 500 502 503 504 505 506 508 511 513 520 522 524 525 527 529 530
 [55] 535 541 542 543 544 545 580 599 618 620 625 630 661 662 663 664 665 666
 [73] 667 668 669 670 693 707 720 725 737 740 752 767 768 769 770 772 792 802
 [91] 806 807 808 809 847 850 852 855 862 870 876 878 880 886 890 891 895 896
[109] 899 900 916 935 936 941 945 965 968 969 970 985 988 990 997
> tmp[[2]]
          id sex   wt     ht mage lit died alive  age
10% 120131.9   1  7.4  69.56 20.9   0    0     1 12.5
25% 120368.8   1  9.0  76.40 24.0   0    0     2 22.5
50% 360111.5   1 11.1  84.80 28.0   0    0     4 37.5
75% 360591.2   2 13.2  94.30 34.0   0    1     6 52.5
90% 360923.0   2 15.2 101.70 35.0   0    2     8 62.5

> table(tmp$newfram[,"age"])

   0  2.5  7.5 12.5 17.5 22.5 27.5 32.5 37.5 42.5 47.5 52.5 57.5 62.5 67.5 72.5 
   1   23   40   68   96   81   92   76   76   73   74   83   78   68   49   20 
77.5 
   2 

# (b) Part 2
> Titanic3 = read.csv("http://biostat.mc.vanderbilt.edu/wiki/pub/Main/DataSets/titanic3.csv",
               stringsAsFactors=F)
> tmp[[4]]
[1] "age"  "fare"
> length(tmp$Rows)
[1] 264
> tmp$Quant
    pclass survived  age sibsp parch     fare
10%      1        0 12.5     0     0  7.56750
25%      2        0 22.5     0     0  7.89580
50%      3        0 27.5     0     0 14.45420
75%      3        1 37.5     1     0 31.27500
90%      3        1 47.5     1     2 78.05082

> table(tmp$newfram[,"age"])

 2.5  7.5 12.5 17.5 22.5 27.5 32.5 37.5 42.5 47.5 52.5 57.5 62.5 67.5 72.5 77.5 
  56   30   29  133  195  166  115   95   72   60   36   26   23    4    4    2