Skip to content

hw3.md

Latest commit

 

History

History
100 lines (72 loc) · 2.67 KB

hw3.md

File metadata and controls

100 lines (72 loc) · 2.67 KB

home | copyright ©2016, tim@menzies.us

overview | syllabus | src | submit | chat

Homework3

Extend Your Table Code

Normalize Numbers

Extend your Table reader (from last week) with methods to "normalize" numbers and return distances between numbers

class Num
 ...
 def norm(i,x):
    tmp= (x - i.lo) / (i.up - i.lo + 10**-32)
    if tmp > 1: return 1
    elif tmp < 0: return 0
    else: return tmp
  def dist(i,x,y):
    return i.norm(x) - i.norm(y)
  def furthest(i,x) :
    return i.up if x <(i.up-i.lo)/2 else i.lo

Normalize Symbols

Extend your Sym class (from last week) to "normalize" symbols (actually, to do nothing at all)

class Sym
  ...
  def norm(i,x)   : return x
  def dist(i,x,y) : return 0 if x==y else 1
  def furthest(i,x): return "SoMEcrazyTHing"

Implement distance between two rows

Implement Aha's algorithm (http://goo.gl/ZspOeL, p42, first para) for the distance between two rows in a table of data (where the columns are symbols or numbers)

For what its worth, here's that code from one of my tools:

  • missing values are denoted "?"
  • my i.cols headers contain the column numbers of the headers. So r1[col.pos] means "ask my column header where to look in the row".
  • All my i.cols Nums or Syms so i can defer to them to find the distances
  • I divide Aha's term by n^1/2 (where n is th e number of variables) so that my distances range 0 to 1.
 UNKNOWN = "?"
 def distance(i,r1,r2,f=2):
    d,n = 0, 10**-32
    for col in i.cols:
      x, y  = r1[col.pos], r2[col.pos]
      if x is UNKNOWN and y is UNKNOWN:
        continue
      if x is UNKNOWN: x=col.my.furthest(y)
      if y is UNKNOWN: y=col.my.furthest(x)
      n    += 1
      inc   = col.dist(x,y)**f
      d    += inc
    return (d**(1/f)) / (n**(1/f))

Test

For data/weather, for the first two rows, find the nearest and furthest rows

Important note: in the distance calc, do not include the dependent variable (the class). Distance is usually computed between independent variables.

Integrated into Ninja.rc

Using the tricks explored last week.

What to hand in

Code

Example output from Test. Make you print out rows1,2 and, for each, print the row closest and furthest away.