Algorithm HajSeyedJavadi

www.Bshams.
ir
a a
1384
www.Bshams.ir
www.Bshams.ir
a
a
a
1
1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1
1
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . big-O 1.1
2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . big
2.1
..........................................................
3.1
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . small o
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . small
4.1
5.1
3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.1
5
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.1
2
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
.2
1.2
10 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1.2
14
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
26
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (Master Theorem) 3.2
2.2
35 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2
37 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2
37
. . . . . . . . . . . . . . . . . Selection Sort
37 . . . . . . . . . . . . . . . . . . . . . . Bubble
www.Bshams.ir
1.5.2
Sort
2.5.2
www.Bshams.ir
38
. . . . . . . . . . . . . . . . . . . Insertion Sort
39
. . . . . . . . . . . . . . . . . . . . Pigeon hole Sort
41 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Binary
42 . . . . . . . . . . . . .Binary
3.5.2
4.5.2
Search
Insertion Sort
5.5.2
6.5.2
43 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Shell
Sort
7.5.2
45 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bucket
Sort1
8.5.2
45 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bucket
Sort2
9.5.2
46
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bin Sort
47 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Counting
47 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Sort
10.5.2
11.5.2
Radix sort
48
. . . . . . . . . . . . . . . . . .
50
. . . . . . . . . . . . . . . . . . . . . . . . . Catalan Number
( )
Trace
12.5.2
6.2
7.2
3
55 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3
56
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (Sparse array) 1.3
57 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3
59
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . max heap
62 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ( )
Binomial Heap 4.3
63 . . . . . . . . . . . . . . . . . . . . . . . . . ( )Binomial
63 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
64
Tree 1.4.3
Max Binomial Tree 2.4.3
. . . . . . . . . . . . . . . . . . . . . . . The Merge Of Max Binomial Trees 3.4.3
64 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
65 . . . . . . . . . . . . . . . . . . . . . . . . .Min
Binomial Heap
Binomial Heap
66 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.4.3
5.4.3
Max Binomial Heap
67 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . FIBONACCI
67 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fibonacci
www.Bshams.ir
3.3
6.4.3
HEAP 5.3
Tree 1.5.3
www.Bshams.ir
68 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Max Fibonacci Tree 2.5.3
68 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fibonacci
68 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Max
Heap
3.5.3
Fibonacci Heap
4.5.3
69 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-3 6.3
72 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-3 1.6.3
73 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-3 2.6.3
77
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-3
83
. . . . . . . . . . . . . . 2-3
84
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Red-Black -
3.6.3
4.6.3
7.3
84 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . - 1.7.3
85
...................................
2.7.3
86 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.7.3
87 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.7.3
89 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.7.3
94 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (Disjoin
sets) 8.3
4
97 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4
98
. . . . . . . . . . . . . . . . . . . . (Greedy Algorithms)
1.4
98 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1.4
101
. . . . . . . . . . . . . . . . . . . MST
2.1.4
109 . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Knapsack
110 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
112 . . . . . . . . . .(timetable
or scheduling)
DIJKSTRA
4.1.4
5.1.4
117 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (devide
www.Bshams.ir
3.1.4
and conquer)
2.4
117
........................................
119
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . merge sort
1.2.4
2.2.4
www.Bshams.ir
121
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Quick Sort
126
. . . . . . . . . . . . . . . . (
) 4.2.4
128 . . . . . . . . . . . . . . . . . . . . . . Dynamic
Programming
128 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
130
3.2.4
....................................
131 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.4

n
k
1.3.4
b 2.3.4
{0, 1}
b 3.3.4
131
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Floyd
133
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.3.4
5.3.4
135 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.3.4
138
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
7.3.4
140 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . b 8.3.4
142 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . b 9.3.4
143
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
144 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B
4.4
&T ( Back Tracking )
5.4
144
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . nb 1.5.4
146
..................................
b 2.5.4
147 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . m-coloring b 3.5.4

148 . . . . . . . . . . . . . . . . . . . . . . . . . . .
Branch and Bound (B&B) 6.4

5
149 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Exploring
149 . . . . . . . . . . . . . . . . . . . . . . . . . . . (Depth
graphs
First Search) DFS
.5
1.5
152 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1.5
www.Bshams.ir
152
. . . . . . . . . . . . . . . . . . . . . . . . . . (Breath First Search) BFS
153
. . . . . . . . . . . . . . . . . . . . . . . . Topological Sort
154
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bellman Ford
4.5
155
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . DAG
5.5
2.5
3.5
www.Bshams.ir
6
157 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . LOOP
173 . . . . . . . . . . . . . . . . . . . . . . . . . (Amortized
Analysis) 2.6
173 . . . . . . . . . . . . . . . . . . . . . . . . . (Aggregate
Analysis)
176
. . . . . . . . . . . . . . . . . . . . . . . . (Accounting Method)
178 . . . . . . . . . . . . . . . . . . . . . . . . . . (Potential
www.Bshams.ir
INVARIANT1.6
Method)
1.2.6
2.2.6
3.2.6
www.Bshams.ir
,

1384
.

1382 ,
,
.

a

1384
www.Bshams.ir
www.Bshams.ir
.1

.
big-O
:
1.1
f : N R+
O(f (n)) = {g:N R+ |c R+ , n N g(n) cf (n)}
f (n) =
3
n
n<1000
2n 2
n100
O(n2 ) = {n2 , n lg n, f (n), ....}

6
f1
f2 O(f1 )
...
...
...
...
.
M
www.Bshams.ir
O(f1 )
f2
-
f2
www.Bshams.ir
.1
6
f2
f1 O(f2 )
f1
f2 O(f1 )
f2 O(f1 )
f1
c
. ,

big-O ,
big-O
.
.
big
:
2.1
f : N R+
(f (n)) = {g : N R+ |d > 0 n N f (n) dg(n)}

:
f (n) < dg(n) = f (n) dg(n)

f (n) dg(n) = f (n) < d0 g(n)
:
f2 (n) (f1 (n)) f1 (n) O(f2 (n))
(f (n)) = O(f (n)) (f (n))
3.1
f : N R+
:
O(n2 ) = {n2 , n2 + 4, nlogn, ...}

(n2 ) = {2n2 , 2n2 +
n, n3 + n2 , n4 , n2 logn, ...}
(n2 ) = {n2 , n2 + n, 2n2 , ...}

:
,
www.Bshams.ir
www.Bshams.ir
SM ALL O .4.1
small o
:
4.1
f : N R+
o(f (n)) = {g : N R+ | c > 0 n N g(n) cf (n)}
n o(2n) 2n o(n):
2n cn
c=1
F = 2n
/ o(n)
n o(2n) c > 0 n N n c2n
c= 13
F = n
/ o(2n)
2n o(n) c > 0
nN
small
:
5.1
f : N R+
(f (n)) = {g : N R+ | c > 0 n N g(n) cf (n)}
(f (n))
: :
o(f (n)) =
M1
d0 > 0 g(n) (f (n))
o(f (n))
:
:
n M1 g(n) d0 f (n) (g(n) (f (n)))

c0 = d20
n M2 g(n) c0 f (n) (g(n) o(f (n))) = g(n) d20 f (n) =
:
M2
2 f (n) d0 f (n) = d0 2d0 = d0 0
d0
n M ax{M1 , M2 }
. ,
6.1
1)f (n) + g(n) O(M AX{f (n), g(n)})

2)f (n) + g(n) (M AX{f (n), g(n)})
: 1
f (n) M AX{f (n), g(n)}
g(n) M AX{f (n), g(n)} =
f (n)+g(n) 2M AX{f (n), g(n)} = f (n)+g(n) O(M AX{f (n), g(n)})
www.Bshams.ir
www.Bshams.ir
.1
: 2
M AX{f (n), g(n)} f (n) + g(n) = f (n) + g(n) (M AX{f (n), g(n)})
f (n) + g(n) O(M AX{f (n), g(n)}), f (n) + g(n) (M AX{f (n), g(n)})
= f (n) + g(n) (M AX{f (n), g(n)})
:
.
log n! (n log n)
:
:
2n)
n! = ( ne )n 2n = log n! = log( ne )n + log 2n
= log n! = n log ne + log 2n
= log n! = n log n log en + log 2n
(n!
( ne )n
= log n! = n log n + log e2nn
= log n! (n log n)
:
.log
n! (n log n)
log n! O(nlog n)
logn! = logn + log(n 1) + . . . + log 1 logn + logn + . . . + logn nlogn

{z
}
|
n
= logn! O(nlogn)
log n! = log n + log(n 1) + . . . + log 1 n2 log n2 =
log n! n2 log n n2 log 2 14 n log n = logn! (nlogn)
= log n! (n log n)
:
:
(n!)2 = (1 2 . . . (n 1)n)2
n
Q
(n!)2 = (1 2 . . . (n 1)n)(n(n 1) . . . 1) =
x(n x + 1)
x=1
y = x + (n + 1)x
(n+1)2
1 = y
x = n+
max =
2
x = 1 y = n, x = n 4 y = n = y
min = n
www.Bshams.ir
www.Bshams.ir
.7.1
n
Y
n (n!)2
x=1
n
Y
(n + 1)2
x=1
nlogn 2logn! 2nlog
= nn (n!)2 (
n+1
2nlogn =
n + 1 2n
) =
logn logn! nlogn
= log n! (n log n)
o(f (n))
7.1
o(f (n)) O(f (n))\(f (n))

o(f (n)) O(f (n))\(f (n))
:

g(n) O(f (n))
g(n) O(f (n))
g(n) o(f (n))
g(n) O(f (n))
g(n) (f (n))
g(n) 5f (n)
d0 0
M1
d0
g(n)
/ (f (n))
n M1 , M1
d = d0
g(n) o(f (n))
c = 2
n M AX{M1 , M2 } g(n) d20 f (n)
c=
g(n)
/ (f (n))
M2
d0 f (n) g(n), n M1
n M2
g(n) o(f (n))
(A\B
= A\(A B)):
o(f (n)) O(f (n))\(f (n)) = O(f (n))\(O(f (n))) (f (n))

= o(f (n)) O(f (n))\(f (n))
. :
:
,
:
www.Bshams.ir
www.Bshams.ir
.1
12
g(n)= n
n
n
:
M = 12 , n 13 g(n) n = g(n) O(n)

: ,
g(n) (n) c >
g(n) (n)
n g(n) cn c >
( )
g(n)
c
nF
= g(n) 6 (n)
g(n) o(n)
g(n) o(n) c > 0
g(n) cn, c = 12 = g(n) n2 F
= g(n) 6 o(n)
.
:
.
:
:
f (n) O(f (n)) c > 0
n N f (n) 1 f (n)
f (n) O(h(n))
g(n) O(f (n))
g(n) O(f (n))
g(n) c1 f (n): n M1

M2 >
M1 > 0
c1 0
0 c2 0 f (n) O(h(n))
f (n) c2 h(n):
:
g(n) c1 f (n)
f (n) c2 h(n)
nM
n M2
M = M AX{M1 , M2 }
= g(n) c1 f (n) c1 c2 h(n) = g(n) ch(n)
= g(n) O(h(n))
.
www.Bshams.ir
www.Bshams.ir
.7.1
n2 O(n)
n2
/ O(n)
n M1 , M1
n O(n2 )
c > 0
n2 cn n c (. )
.
:
:
f (n) O(f (n))
O,
, f (n) (f (n)) f (n) (f (n))

.
g(n) (f (n))
g(n) (f (n)) = g(n) O(f (n)), g(n) (f (n)) = f (n) (g(n)),

f (n) O(g(n)) = f (n) (g(n)) Og(n)) = f (n) (g(n))
.
f (n) (g(n))
O,
, g(n) (h(n)) =
f (n) O(g(n))
, g(n) O(h(n)) f (n) O(h(n))
f (n) (g(n))
, g(n) (h(n)) f (n) (h(n))
= f (n) (h(n))
: :
L=
(n)
limn fg(n)
,
f, g : N R
+
:
.f (n)
.g(n)
.g(n)
o(g(n))
o(f (n))
(f (n))
L = 0 (1
L = (2
0 < L < (3
:
www.Bshams.ir
www.Bshams.ir
.1

. O, o, ,
:
:
f (n) O(g(n)) = 2f (n) O(2g(n) )

:
n log2
n logn2 O(logn!
O(2log2 ) = nn O(n!)
2 ) = 2
n
n!
2
log n
2
2 logn2 O(logn2 ) = 2
O(2log2 ) = n2 O(n)
n
f (n) O(g(n)) = (f (n))k O(g(n)k )

:

g f
(.
f (n) cg(n) = (f (n))k ck (g(n)k ) = (f (n))k O(g(n)k )
f (n) O(f 2 (n))

:
f (n) = n1 = f 2 (n) = n12 = f (n) 6 O(f 2 (n))
www.Bshams.ir
www.Bshams.ir
.2

1.2

.

:
a0 tn + a1 tn1 + + ak tnk = f (n) ai R, f : N R0 , k N

n
k ,
t0 , , tk1
. .
.
,f (n)
0
:
a0 tn + a1 tn1 + + ak tnk = 0
xi
ti .

: .
www.Bshams.ir
www.Bshams.ir
.2
10
a0 xn +a1 xn1 + +ak xnk = 0 xnk (a0 xk + a1 xk1 + + ak ) = 0

|
{z
}
()
.
r1 , . . . , rk
1.1.2
k (1
tn = c1 (r1 )n + c2 (r2 )n + . . . + ck (rk )n

, (2

(x rp )m )
: ((x
rp )
m+1
rp
tn = c1 (r1 )n + c2 (r2 )n + . . . + cp0 (rp )n + cp1 n(rp )n +

. . . + cpm1 nm1 (rp )n + . . . + ct (rt )n
, (3
f
.
:
:
g(n) =
5g(n 1) 6g(n 2)
n1
else
:
g(n) 5g(n 1) + 6g(n 2) = 0 = x2 5x + 6 = 0
x1 = 3, x2 = 2 g(n) = c1 3n + c2 2n
g(0) = 0, g(1) = 1 c1 = 1 , c2 = 1 = g(n) = 3n 2n
www.Bshams.ir
www.Bshams.ir
.1.2
11
2tn1 tn2
tn = t0 = 1
t1 = 3
:
tn 2tn1 + tn2 = 0 = x2 2x + 1 = 0 x1 = x2 = 1
t0 = 1
tn = c1 + c2 n
, t1 = 3 c1 = 1, c2 = 2 = tn = 1 + 2n
2tn1 + 1
tn = t0 = 0
t1 = 1
:
tn = 2tn1 + 1
, tn1 = 2tn2 + 1 tn tn1 = 2tn1 2tn2
tn 3tn1 + 2tn2 = 0 x2 3x + 2 = 0
x1 = 1, x2 = 2
, t0 = 0, t1 = 1 = tn = 2n 1
:
a0 tn + a1 tn1 + a2 tn2 + . . . + ak tnk =

bn1 p1 (n) + bn2 p2 (n) + . . . + bnm pm (n)
( .
ai , bj
di pi (n)
(a0 xk + a1 xk1 + + ak )(x b1 )d1 +1 (x b2 )d2 +1 (x bm )dm +1 = 0

.
f
.
:
tn 7tn1 + 12tn2 = 2n + n + 3n (n + 1)
:
(x2 7x + 12)(x 2)0+1 (x 1)1+1 (x 3)1+1 = 0

(x 3)(x 4)(x 2)(x 1)2 (x 3)2 = 0
www.Bshams.ir
www.Bshams.ir
.2
12
r1 = 1, r2 = 1, r3 = 2, r4 = 3, r5 = 3, r6 = 3, r7 = 4
tn = c1 (1)n + c2 n(1)n + c3 (2)n + c4 (3)n + c5 n(3)n + c6 n2 (3)n + c7 (4)n

:
:
t(n) =
1
2t( n2 ) + n
n=1
o.w
:
.
n = 2k t(n) = t(2k )
t(2k ) = 2t( 22 ) + 2k = 2t(2k1 ) + 2k

k
g(k) = t(n)
g(k) = 2g(k 1) + 2k (x 2)(x 2) = 0 r1 = 2, r2 = 2
= g(k) = c1 2k + c2 k 2k
2k = n = log2 n = k
t(n) = c1 n + c2 n logn2
t(1) = c1 = 1
t(2) = 2t(1) + 2 = 4 = 2c1 + 2c2 c2 = 1

= t(n) = n + n logn2
:
:
T (n) =
3
2
n
1 n
1
2T(2) 2T(4) n
n=1
n=2
o.w
:
:
n = 2k = T (2k ) = 32 T (2k1 ) 12 T (2k2 ) 21k =
g(k) = 32 g(k 1) 12 g(k 2) ( 12 )k = (x2 32 x 12 )(x 12 ) =

r1 = r2 = 12 , r3 = 1 = g(k) = c1 + c2 ( 12 )k + c3 k( 12 )k =
www.Bshams.ir
www.Bshams.ir
.1.2
13
T (n) = c1 + c2 n1 + c3 lgnn
= T (n) = 1 +
c1 + c2 = 1
, c1 + c22 + c23 = 32
c1 + c42 + c23 = 32
lg n
n
. :
t(n) + nt(n 1) = 2n!

t(0) = 1
:
t(n) + nt(n 1) = 2n! =

t(n)
n!
t(n)
n!
t(n1)
(n1)!
=2
= g(n)
g(n) + g(n 1) = 2 = (x 1)(x + 1) = 0
= r1 = 1, r2 = 1 = g(n) = c1 (1)n + c2 (1)n

t(n) = 21 n!(1)n + 32 n!
T (n) =
n=0
n=1
1+T (n1)
T (n2)
o.w
:
T (0) = a
T (1) = b T (2) = 1+b

a
T (3) = 1+a+b
ab
T (4) =
T (5) = a T (6) = b T (7) = 1+b

a = T (n) = T (n mod
a+1
b
5)
n>4
:
www.Bshams.ir
www.Bshams.ir
t(n) =
.2
1
4t(n
1)
14
n>1
n=1
:
3 , t(4) = 11 = t =
t(2) = 13 , t(3) = 11
n
41
pn+1
qn+1
pn
qn
n
= t(n + 1) = 4qnqp
= qn+1 = 4qn pn
n
pn+1 = qn = pn = qn1
qn+1 = 4qn qn1 = (x2 4x + 1) = 0 = x = 2 3
= qn = c1 (2 + 3)n + c2 (2 3)n q2 = 3, q3 = 11
2.2

,
b

b
.
www.Bshams.ir
www.Bshams.ir
.2.2
15
: b
, 321

i + j + x = 1 + 2 + 3 x = 6 (i + j)
i j
6-(i+j)
void hanoi(int n,int i,int j)

{
if(n > 0) {
hanoi(n-1,i,6-(i+j)) ;
cout i j ;
hanoi(n-1,6-(i+j),j);
}
}
T (0) = 0
T (n) = T (n 1) + 1 + T (n 1) =
T (n) = 2 T (n 1) + 1 T (n) = 2n 1
www.Bshams.ir
www.Bshams.ir
.2
16
:
b
)
(.
Void Hanoi (int i , int j)

{
if (n > 0)
{
Hanoi(n-1, i , j);
cout i 6 (i + j);
Hanoi(n-1, j , i);
cout 6 (i + j) j;
Hanoi(n-1, i , j);
}
}
T ( 0) = 0
T ( 1) = 2
T (n) = 3 T (n 1) + 2
T (n) = 3n 1
.

.

).
(.
www.Bshams.ir
www.Bshams.ir
.2.2
17
a
a[1] a[2] .............
a[n]
a[3]
a[3]
a[3]
a[3]
a[2]
a[1]
a[2]
Void Coloring (int n){

if (n == 1){
printf();
printf();}
else if (n == 2) {
printf();
printf();
printf();
printf(); }
else if (n == 3) {
printf();
printf();
printf( );
printf( );
else {
.Coloring(n-1);
.Coloring(n-2);
.Coloring(n-3);
}
}
www.Bshams.ir
printf();
printf();
printf( ); }
www.Bshams.ir
.2
18
4
a(n) =
7
1 a(n 1) + 1 a(n 2) + 1 a(n 3)
n=1
n=2
n=3
else
:
.
.
:

n-2
.
n-2
f .
| {z } |
{z
n2
}
:
T (n) =
T (n 2) + 3
n-1
n=2
n-1
n = 2k
= T (n) = 32 n 2
( 32 (n 1) 2) + 2f ,
n-1
: .
2T ( n2 ) , n ) 1
. T (n) = 2T ( n
2 ) + 2 .
(. T (n) = 32 n 2

www.Bshams.ir
www.Bshams.ir
.2.2
19
T (n) =
n=1
3 (n 1)
n = 2k + 1
2n 2
n = 2k
: AVL
O(log n)
log n
.
.
:
.
:
4
AVL
n-1
AVL
T(1)
T(0)
T(2)
2
T (n) =
4
T (n 1) + T (n 2) + 1
www.Bshams.ir
T(3)
n=0
n=1
n=2
o.w
n-2
T(n-1)
T(n-2)
T(n)
www.Bshams.ir
.2
20
T (n) = T (n 2) + T (n 1) + 1
1 , 1 , 2 , 3 , 5 , 8 , 13 , 21 , 34 , 55

1 , 2 , 4 , 7 , 12 , 20 , 33 , 54
T (n) = f (n + 2) + (1)
f (n) = f (n 1) + f (n 2)
f (0) = 1 , f (1) = 1
f (n) = f (n 1) + f (n 2)
f (0) = 0, f (1) = 1
x2 x 1 = 0
r1 = 1+2 5
f (n) = c1 ( 1+2 5 )n + c2 ( 12 5 )n
f (n) = 15 ( 1+2 5 )n 15 ( 12 5 )n
AVL
, r2 = 12 5
1 1 + 5 h+2
n T (h) = f (h + 2) 1 = n (
)
1 =
5 2
1
1
1
+ 5 h+2
1) log 5 +(h + 2) =
n+1 (
)
= log(n+
5 2
1) log 5 = h O(log 5(n+1) ) = h O(logn )

h + 2 log(n+
log2 + log 5 = 1.44 log2
(n+1)
0.33 = h O(logn
2)
= h 1.44 log2
logn2
(n+1)
1)
h + 2 log(n+
log 5 = log2
(n+1)
+1.67
www.Bshams.ir
www.Bshams.ir
.2.2
21
n
.
(c1
< c2 < < cn
n)
ck
ck+1 < ... < cn
c1 < c2 < ... < ck1

T (k 1)
T (1) = 1
T (n) =
Pn
k=1
T (n k)
T (2) = 2
1 2n
T (k 1) T (n k) = n+
1 n
( )
:
A1 , A2 , , An
?
:
T ( 0) = 0
T ( 1) = 1
(A)
T ( 2) = 1
(AB)
T ( 3) = 2
T ( 4) = 5
..
.
A(BC), (AB)C
A(BCD), (AB)(CD), (ABC)D
T (4) = T (1).T (3) + T (2)T (2) + T (3).T (1)
(A1 Am ) (Am+1 An )
{z
}|
|
{z
}
m
T (n) =
n
X1
m=1
nm
T (m)T (n m) =
1 2n 2
n n1

:
.
:
www.Bshams.ir
www.Bshams.ir
.2
22
10
n
n-1
= a(n)
= a(n 1)
(. )
a(n 1)
}|
{z
}|
1,2,...,9
.
n-1
b(n 1)
(. )
b(n 1)
}|
= a(n) = 9a(n 1) + b(n 1)

1
10
10
n1
.........
10
(n-1)
= a(n-1)+b(n-1)=10n1
a(n 1) + b(n 1) = 10n1 b(n 1) = 10n1 a(n 1)
= a(n) = 9a(n 1) + 10n1 a(n 1) = 8a(n 1) + 10n1
www.Bshams.ir
www.Bshams.ir
.2.2
23
1
T(1)=2
2
1
3
T(2)=4
3
5
1
2
T(3)=7
T (1) = 2, T (2) = 4, T (3) = 7, T (4) = 11, . . .

n .

T (n 1)
T (n) = T (n 1) + n = T (n) =
n 1
T (n) = T (n 1) + n
n(n + 1)
n 1
+1
n 2
? 2
www.Bshams.ir
www.Bshams.ir
.2
24
T ( 2) = 1 :
2
T ( 3) = 1 :
3
T ( 4) = 2 :
2 + 2, 4
T ( 5) = 3 :
2 + 3, 3 + 2, 5
T ( 6) = 5 :
2 + 2 + 2, 2 + 4, 3 + 3, 4 + 2, 6
" 2+ ' k 2 k
: , k 1
T (5) :
{2 + (3)}, {(2 + 1) + 2}, {(4 + 1)}

|
{z
} | {z }
3+2
T ( 6) :
{2+(2+2)}, {2+(4)}, {(2 + 1) + 3}, {(3 + 1) + 2}, {(5 + 1)}

|
{z
} |
{z
} | {z }
3+3
T(n)=
4+2
n=2
n=3
T (n 1) + T (n 2) o.w
:
0,1,2
,
.
:
.n :
.
.
.
an
bn
cn
n = an + bn + cn
an = bn1 + cn1
, bn = an1 + cn1
, cn = an1 + bn1 + cn1
cn = n1 cn1 = n2
= n = an + bn + cn = 2an1 + 2bn1 + 3cn1 =
www.Bshams.ir
www.Bshams.ir
.2.2
25
2(an1 + bn1 + cn1 ) + cn1 = 2n1 + cn1

= n = 2n1 + n2
:
.
nm
.
.
A b
T(n,m-1)
T (n, m 1)
{z
T (n 1, m)
B
d - T(n-1,m)

: .
T (m, n) = 1 T (n, m 1) + T (n 1, m) 1 , T (n, 0) = T (0, m) = 0

m+n
T (m, n) =
n, m
:

,
.
:
.
n*n
www.Bshams.ir
www.Bshams.ir
.2
26
.
?
:
.
?
:
.
, :
0 1
.
{}
{4}
{3}
n-1
n2

.
:
tic tac toe
(Master Theorem)
f : N R+ , b > 1 , a 1
3.2
T (n) = a T ( nb ) + f (n)
www.Bshams.ir
>
f (n)
O(nlog
a
b
(1
www.Bshams.ir
(MASTER THEOREM) .3.2
27
.T (n)
T (n) (nlog b (log n)k+1 )

a
0
.
< <
, >
T (n) (f (n))
f ( nb )
f (n) (nlog b (log n)k )
(nlog b )
f (n) (nlog
< f (n)
b nb c
d nb e
a
b +
(2
(3
n
b )
( .
:
.

T(n)
n = 1, b, b2 , ..
b > 1
T(n)
T(n)=
n
n2
T(n)
T(n)O(n)
n=1,2,3,..
otherwise
T(n)O(n2 )
:
.
T (n) = aT ( nb ) + f (n)
www.Bshams.ir
www.Bshams.ir
.2
28
b>
1 n
b . .
. b

b n
. ,
: 1
f (n)
b>
1 a 1
T(n)
T(n)=
(1)
n=1
aT ( nb ) + f (n)
n = bi
1
logn
b
T (n) = (n
loga
b
)+
(1)
aj f (n/bj )
j=0
:
f (n)
.
f (n/b)
a2
T (1) = (1)
loga
b
f (n/b2 )
.
j
n
logb .
f ( nb )
lognb
.
www.Bshams.ir
www.Bshams.ir
29
- f(n)
f(n)
a
6

f(n/b)
a
j
f(n/b)
a
?
f(n/b)
a
lognb f (n/b2 ) ...................................................................

a
a
a
a
a
a
a
.. .. .. .. ..
.. .. .. .. ..
.. .. .. .. ..
?
(1)(1)(1)
|
..
..
..
..
..
..
f (n/b2 ) a2 f (n/b2 )
a
a
.. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. ..
.. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. ..
.. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. ..
a
(nlogb )
(1)(1)(1)(1)(1)(1)(1)
{z
}
..
..
..
nlogb
af(n/b)
(1)
, .
aj f (n/bj )
logb n1
aj f (n/bj )
j=0

(1)
(nlogb )

:
, (1
.
. f (2
(3
.
: 2
f (n)
www.Bshams.ir
b>
g(n)
1 a 1
www.Bshams.ir
.2
30
1
logn
b
g(n) =
(2)
aj f (n/bj )
j=0
,
>
f (n) = O(nlogb )
.
nb
g(n) = O(n
(1
loga
b
g(n) = (nlogb . log n) f (n) = (nlogb )(2

c<
.g(n)
1 (3
= (f (n))
af (n/b) cf (n)
f (n) = O(nlogb )
j loga
b
n
(2) f ( j )
b
= O((n/b )
)
:
g(n) = O
1
logn
b
aj
j=0
n logab
bj
(3)
1
logn
b
j=0
n logab
bj
1
logn
b
=n
X ab j
a
blogb
loga
b
j=0
1
logn
b
=n
loga
b
(b )
j=0
=n
loga
b
= nlogb
b logb 1
b 1
n 1
b 1
.
.
www.Bshams.ir
b
loga
b
O(n ) = O(nlogb )
www.Bshams.ir
31
: (3)
a
g(n) = O(nlogb )
.
f (n) = (nlogb )
f (n/bj ) = ((n/bj )logb )

: (4) (2)
1
logn
b
g(n) =
aj
n logab
j=0
bj
(4)
1
logn
b
j=0
n logab
bj
=n
1
logn

b
loga
b
blogb
j=0
1
logn
b
=n
loga
b
j=0
j
= nlogb lognb
: (4)
g(n) = (nlogb lognb ) =

a
g(n) = (nlogb log n)

.
(2)

af (n/b) cf (n)
.
f (n/b) (c/a)f (n) n b
g(n) = (f (n))
c<1
f (n/b) (c/a)f (n) f (n/bj ) (c/a)j f (n) aj f (n/bj ) (cj f (n))
www.Bshams.ir
www.Bshams.ir
.2
32
: (2)
1
logn
b
g(n) =
aj f (n/bj )
j=0
1
logn
b
cj f (n)
j=0
f (n)
cj
j=0
= f (n)
1C
= O(f (n)).
,

.g(n)
= (f (n))
g(n) = (f (n)), g(n) = O(f (n))

. 2
f (n)
b
b > 1 a 1
T(n)
T(n)=
(1)
n=1
aT ( nb ) + f (n)
n = bi
: .
>
f (n) = O(nlogb )
.T (n)
.T (n)
www.Bshams.ir
= (nlogb log n))
= (n
loga
b
(1
)
a
f (n) = (nlogb ) (2
www.Bshams.ir
33
0 , f (n) = (nlog + ) (3
af (n/b) cf (n) n c < 1

a
b
>
.T (n)
= (f (n))
:
: 1 (1) 2
:
T (n) = (nlogb ) + O(nlogb ) = (nlogb ),

: 2
a
T (n) = (nlogb ) + (nlogb log n) = (nlogb log n),

: 3
a
T (n) = (nlogb ) + (f (n)) = (f (n)),

:
f (n) = (n
www.Bshams.ir
loga+
b
www.Bshams.ir
.2
T(n)=
34
. :
3 T ( n2 ) + n
n=1
else
( ):
f (n) =
n=
T (n)
T(n)=
1
n2
3
O(nlog 2 )
3
(nlog 2 )
0
2T ( n2 ) + n 1
a=3, b=2
. :
n=1
else
( ):
T (n) = 2 T ( 2 ) + n 1
f (n) = n 1 (n)
n
(n) =
2
(nlog 2 (log
a=2, b=2
n)0 ) T (n) (n(log n)0+1 )
T (n) (n log n)
T (n) =
. :
1
3 T ( n2 ) + n2
n=1
else
( ):
3
(nlog 2 + )
f (n) = n2
a
= 3 ( 2 )2 n2 = 45
T (n) (n2 )
f ( nb )
T (n) =
a=3, b=2
. :
2 T ( n2 ) + log n!
n=1
else
:
f (n) = log n! (n logn)

(n logn) =
2
(nlog 2 (logn))
a=b=2
T (n) (n (logn)1+1 ) T (n) (n (logn)2 )
www.Bshams.ir
www.Bshams.ir
.4.2
35
4.2

. ,

t2
t1
b .t1
(t2 )
t2
t1
p2 , p1
, ,
.
t1 + t2
p1 , p2
t1 + t2 (max{t1 , t2 })

(1)
(1)
:while
i1
while (i <= m)
p(i)
ii+1
,
1:
i1
O((m + 1) 1) +
O(m) +
L O(1) +
O(mt) +
p(i)
i i + 1
1
).p(i) O(t) . t, p(i)
(
www.Bshams.ir
www.Bshams.ir
O(m)
.2
36
while goto
L < O(2 + 3m + mt) L O(max{1, m, mt})

.

ti
p(i)
p(i)
L O max{1, m,
m
X
ti }
i=1
:for
while
for
while for
for i 1 to n 1 do
for j i + 1 to n do
write(*);
ti =
n
X
j=i+1
n
X1
i=1
ti =
(1) = n (i + 1) + 1 = n i
n
X1
i=1
(n i) = n(n 1)
(n 1)n
(n 1)n
O(n2 )
,
.
n2 n
1 2 ................. n
1 .*....*.
*
*......
2.
...
*....
....
...
....
...
....
...
....
...
.*...
.
....
n
*.
n2 n
www.Bshams.ir
O(n2 )
O(1)
www.Bshams.ir
.5.2
37
Selection Sort
5.2
1.5.2
procedure Selection Sort( T [ 1 n ])

fori 1 to n-1 do

minx T[i]
(1)
O(1)
minj i
for j i+1 to n do
if (T[j] < minx)then

minx T[j]
O(1)
(1)
minj j

T[minj] T[i]
O(1)
(1)
T[i] minx
:
n
n
P1
P
(O(
1
)
+
O(1)) O(n2 )
L
i=1
j=i+1
i=1
j=i+1
n
n
P1
P
((1) +
(1)) (n2 )
L>
L (n2 )
Bubble Sort
2.5.2
Procedere Bubble Sort(T[1..n])

for i 1 to n-1 do
for j i+1 to n do
if(T [j] < T [i]) then
swap(T [i], T [j])
n 1
n(n1)
www.Bshams.ir
(1)
O(n2 )
www.Bshams.ir
.2
38
Insertion Sort
3.5.2
Procedere Insertion Sort(T[1..n])

for i 2 to n do
x T [i]
j i-1
while (j > 0 && T [j] > x)
T[j+1] T[j]
j
T[j+1] x
(n)
n
P
i=2
false
while
(O(1) + O(i)) O(n2 )
.
.
(Best Case Analysis)

,

while
O(1)
while
.
.
(n)
n-1
(n)
O(n)
(Worst Case Analysis)

while
(i)
(n2 )
(i )
O(n2 )
www.Bshams.ir
(n2 )
www.Bshams.ir
.5.2
39
(Average Case Analysis)
E(x) =
i-1
i-2
.....
.....
x f (x)
xX
T[1..i]
ci =
i
X
k=1
f (x) = 1i
ci
i
1=1X
1 i(i + 1) = i + 1
k=
i
i
i
2
2
k=1
n
X
i=2
ci =
n
X
i+1
i=2
sort
n
n
X
1X
1 n(n + 1)
(
i+
1) = (
1 + n 1) = (n2 )
2 i=2 i=2
2
2
insertion sort

.
.
n!
Pigeon hole Sort

u
n!
4.5.2
,
.
.
.
www.Bshams.ir
www.Bshams.ir
.2
40
Procedure pigeon hole sort(T[1..n])

let m = max{T [i]} T [i] Z >0 , 1 i n
array u[1..m]
for i 1 to m do
// (m)
u[i] 0
for i 1 to n do
// (n)
k T[i]
u[k]++
k 1
for i 1 to m do
while u[i] 6= 0 do
T[k] i
u[i]
k++
:
m+1 for
n
P
i=1
(u[i] + 1)
for while
: ,
m
P
(u[i] + 1) =
i=1
m
P
i=1
u[i] +
m
P
i=1
1=n+m
(n + m)
T
u
2
1
0
1
2
0
3
3
0
2 1
4 5
0 0
1
2
2
4
3
2
4
1
5
2
1 1 2
| {z } |
{z
m=5
}| {z }
3
5 5
| {z }
www.Bshams.ir
www.Bshams.ir
.5.2
41
Binary Search
5.5.2
Binary Search (A, temp){

left = 0;
right = lenght(A)-1;
while (left< right) do
middle= (left+right) / 2;
if (temp > A[middle])
left= middle + 1;
else
right= middle-1;
return left;}
.
.
temp
temp
:Binary Search
i lef t, j right,
i = 0, j = n 1
temp
d=j i+1=n10+1 =n
,
if
k = middle = i+j
2
middle, right, lef t k, j, i
j, i
d,
if
right, left
temp A[middle] i = k+ 1, j = j, d = j i+ 1 = j k = j ( i+j

2 )=
ji
ji+1
<
= d d < d
2
ji
else j = k 1, i = i, d = j i + 1 = k 1 i + 1 = i+j
2 i = 2 <
ji+1
= d d < d
www.Bshams.ir
www.Bshams.ir
.2
42
ij j i0j i+11d1
d0 = n d1 < d20 = n2 d2 < d21 = n4
dk 1 2nk 1 n 2k logn2 k
k = dlogn2 e O (logn2 )
Binary Insertion Sort
6.5.2
Binary Insertion Sort(A[1...n]){

for (i = 1; i < n; i + +){
temp= A[i];
left=1;
right=; i
Binary Search algorithm // for array A , temp=A[i], left=1 , right=i
for (j = i; j > lef t; j )
swap(A[j-1],A[j]);
}
}
worst case
n
P1
i=1
insertion sort
O(n2 )
Binary insertion sort
O(n2 )
i+1
blog2 c =
n
P
www.Bshams.ir
n
P
i=2
blog2 c O
average case
O(n2 )
O(n2 )
O(nlog n)
blog2 c ' (n + 1)blog2
c+2
blog(n+1)c+1
or

O(n2 )
n+1
i=2
O(nlogn)
best case
n
P
i=2
log2
Rn
x
2 log2 dx O(nlogn)
+2
www.Bshams.ir
.5.2
43
Shell Sort
7.5.2
void Shell-Sort(int gap,int A[1...n]){

while((gap/=2)1){
for(i=0;i < length(A);i++){
int j=i;
while((j gap) && (A[j-gap] > A[j])){
swap(A[j - gap],A[j])
j -=gap;
}
}
}
}
,
gap
gap
:
4
. 1 4 6 7 ,
2
8 6 , 1 2 ,
5 2

.
www.Bshams.ir
www.Bshams.ir
.2
44
Len=8=gap
2
|
7
|
4
{z
|
8
{z
5
}
{z
6
}
{z
gap=4
10
10
2
7 4
6
1 8 5
| {z } | {z } | {z }
| {z } | {z } | {z }
| {z } | {z }
| {z } | {z }
| {z }
|
{z
}
gap=2
8
1 6 2 7 4
5 10
|{z} |{z}
6|{z} |{z}
2
4
7|{z} |{z}
|{z}
8 |{z}
5
4
6
|{z}
7
|{z} 5
6
5
gap=1
shell-sort
10
gap= 21
while
best case
worst case
n log n
n1.5
average case
n1.25
gap
www.Bshams.ir
www.Bshams.ir
.5.2
45
Bucket Sort1
8.5.2
unsighned const m=max

void Bucket-Sort1(int A[ ],int n)
int buckets[m];
for (int i=0;i<m;i++)
buckets[i]=0;
for(i=0;i< n;i++)
++buckets[A[i]];
for(int j=0,i=0;j< m;j++)
m
P1
for(int k=buckets[j];k>0;- - k)
//
j=1
buckets[j] = n
A[i++]=j;
O(n+m)
Pigeon hole sort
Bucket Sort2

Bucket-Sort
. ,
[0, 1)
Bucket-Sort2(A)
n [A]
for i 1 to n do
insert A[i] into list B[bnA[i]c]
for i 0 to n-1 do
sort list B[i] with insertion sort
concatenate the list B[0],B[1],...,B[n-1] together in order
www.Bshams.ir
9.5.2
www.Bshams.ir
.2
A
1 .78
2 .17
3 .39
4 .26
5 .72
6 .94
7 .21
8 .12
9 .23
10 .68
46
- .17
- .23
- .12
- .21
- .39
- .68
- .72
- .26
- .78
- .94
Bin Sort
for(int i=0;i<m;i++)
10.5.2
//O(m)
bin [i]= -1;

for(int i=0;i < n;i++)
bin[a[i]]=a[i];
//O(n)
j=0;
for(int i=0;i < m;i++){
if(bin[i]!= -1)
//O(n+m)
a[j]=b[i];
j++;}
.
O(m + n)
,
.
a:
3 7 2 8 1
bin:
-1 1 2 3 -1 -1 -1 7 8
a:
www.Bshams.ir
1 2 3 7 8
www.Bshams.ir
.5.2
47
Counting Sort
11.5.2
counting-sort
for i0 to k do
k is maximum of elements
c[i]0
for j 1 to n do
c[a[j]]c[a[j]]+1;
for i 2 to k do
c[i] c[i]+c[i-1];
for j n downto 1 do
b[c[a[j]]] a[j];
c[a[j]] c[a[j]] -1 ;
log n!
(n + k)
dlog n!e
dlog n!e (n log n)

,
(n log n)
(n log n)
Radix sort
stable
12.5.2
. .
sort
(n)
O(n(n))
.
.
www.Bshams.ir
www.Bshams.ir
.2
48
25 13 43 124 313 513 5 1 999 96 205

0
1 - 531 - 1
2
3 - 13 - 43 - 313
4 - 124
5 - 25 - 5 - 205
6 - 96
7
8
9 - 999
531,1,13,43,313,24,25,5,205,96,999
.
( ) Trace
6.2
T, (n) = T ( n5 ) + T ( 710n ) + (n)

T, (1) = C
.
(n)
(n)
O( 710n )
O( n5 )
n
)
O( 25
O( 750n )
O( 750n )
n
O( 49
100 )
.
.
www.Bshams.ir
www.Bshams.ir
( ) TRACE .6.2
49
7
10
i
n 1 = n
T (n)
10
7
dlogn10 e
X7
i=0
53
>
10
= logn 1
7
53
T (n)
i
= logn10 i = i = dlogn10 e =
7
9
10
i
n=
9
10
logn10 +1
7
9
1 10
> logn10 = 3 logn5 > logn10 =

7
dlogn10 e
X7
i=0
9
10
i
3X
logn
5
i=0
9
10
i
T (n) = T ( n3 ) + T ( 23n ) + (n)
(n)
( 23n )
( n3 )
( n9 )
i
2
3
( 29n )
n 1 = n
( 29n )
i
3
2
m
l
= logn3 1 = i = logn3
2
l
= T (n)
logn3
X2
i=0
www.Bshams.ir
( 49n )
m
n O(n logn3 )
2
www.Bshams.ir
.2
50
Catalan Number
7.2
:1 n
. ,
.
1
( ) .
. ,
,

, .
, )
(.
0, n ,bn
0 i n bi .
n
,bn+1
)
.
, .
b0 = 1
www.Bshams.ir
www.Bshams.ir
CATALAN NUMBER .7.2
51
www.Bshams.ir
n,
0 (1)
www.Bshams.ir
.2
52
.b0 bn

b1 bn1
bn+1
n 1,
n + 1,
1 (2)

.
.
.
n i,
bi bni
i (i+1)
bn+1
.
.
.
,
.
bn b0
n (n+1)
bn+1
n 0 ,
bn+1 = b0 bn + b1 bn1 + b2 bn2 + + bn1 b1 + bn b0
X
n=0
bn+1 xn+1 =
(b0 bn + b1 bn1 + + bn1 b1 + bn b0 )xn+1
(1)
n=0
a0 , a1 , a2 ,
a
,
f (x) =
i=0
a0 , a1 ,
ai xi
[f (x)]2
a0 a0 , a0 a1 + a1 a0 , a0 a2 + a1 a1 + a2 a0 ,
, a0 an + a1 an1 + a2 an2 + + an1 a1 + an a0
b0 , b1 , b2 ,
f (x) =
n=0
bn xn
(f (x) b0 ) = x
www.Bshams.ir
X
n=0
(1)
(b0 bn + b1 bn1 + + bn b0 )xn = x[f (x)]2
www.Bshams.ir
CATALAN NUMBER .7.2
53
x[f (x)]2 f (x) + 1 = 0
[1
f (x) =
1 4x = (1 4x)1/2 = 1/2
0
1 4x]
(2x)
1/2
1

(4)n
1/2
n
=
=
=
=
=

1/2 (4x)2 +
(4x) +

n 1, xn
(1/2)((1/2) 1)((1/2) 2) ((1/2) n + 1)

(4)n
n!
(1/2)(1/2)(3/2) ((2n 3)/2)
(4)n
(1)n1
n!
(1)2n (1)(3) (2n 3)
n!
(1)2n (n!)(1)(3) (2n 3)(2n 1)
(n!)(n!)(2n 1)

(1)
(1)(2)(4) (2n)(1)(3) (2n 1)
2n
=
(2n 1)(n!)(n!)
(2n 1) n
, ;
f (x)
"
"
X
1
2n
1
f (x) =
1 1
2x
(2n 1) n
n=1
xn+1
X
n=1
www.Bshams.ir
(2n 1)
2n
n
xn
f (x)
##
xn
xn
bn
bn
www.Bshams.ir
.2
54
bn =
1
1
2 2(n + 1) 1
(1894 1814)

2(n + 1)
(n + 1)
1
(2n)!
=
(n + 1)!(n!)
(n + 1)
bn
2n
. ,
.
.
www.Bshams.ir
b6 = 132
x1 x2 x3 xn
b5 = 42, b4 = 14, b3 = 5, b2 = 2, b1 = 1, b0 = 1
www.Bshams.ir
.3

, f
.
.

.
.

.
char, bool ,int: ,

. ,
struct , union
: .
f :
:
x : array [a..z] of integer

int x[26];
www.Bshams.ir
www.Bshams.ir
.3
56
(Sparse array)
1.3

f ,
.
,
.
.
:
2
3
8
13
11
1 29 U 3
8
2
4
?
z
1
ctr
b[
a[ctr]
ctr
ctr
]
.
ctr=0 ctr=ctr+1=1 T[4]=11 a[ctr]=a[1]=4 b[4]=1


:
1 b[i] ctr
www.Bshams.ir
a[b[i]] = i
www.Bshams.ir
57
ctr
.2.3
b[8]=2
1 b[8] 3
a[b[8]]=a[2]=8
2.3
.
. :
. :
. ,
) :
(Depth)
. (
:(Level)
.
:
a
b
c
d
e
f
2
1
0
0
0
0
0
1
2
2
2
1
2
1
0
0
0
1
www.Bshams.ir
I(T)
E(T).
www.Bshams.ir
.3
T(n)= E(T)-I(T)
58
.
?
h
h+1
?
?
T(n+2)
T(n)
E(T ) = E(T 0 ) h + h + 1 + h + 1
I(T ) = I(T 0 ) + h
E(T ) I(T ) = E(T 0 ) I(T 0 ) + 2 T (n + 2) = T (n) + 2
f :
.
:
) f
.(

f
,
.
.
:

,
www.Bshams.ir
www.Bshams.ir
MAX HEAP
59
.3.3
,
.
2
4
8
15
13 14
11 12
9 10
. ,
max heap
3.3

).
.
heap
(.
2i , 2i+1
, .
(1)
,
.
1 b n2 c
b c + 1n
| 2 {z
}
dn
2e
. (
sift-up
d n2 e )
sift-down (perculate)sift-up, heap
sift-down
www.Bshams.ir
www.Bshams.ir
.3
60
procedure alter heap (T[1..n],i,v)

{ T[1..n] is a heap ,the value of T[i] is set to v and the heap
property is re-established we suppose the 1 i n }
x T[i]
T[i] v
if v < x then sift-down (T,i)
else percolate (T,i)
.....................................................................
procedure sif t down (T[1..n],i)
{ this procedure sifts node i down so as to re-establish the
heap property in T[1..n] we suppose that T would be a heap if T[i]
were sufficiently large we also suppose that 1 i n }
ki
repeat
j k {find the larger of node j }
if (2j n and T[2j]> T[k])
k 2j
if (2j < n and T[2j+1]> T[k])
k 2j+1
exchange T[j] and T[k]
{if j == k then the node has arrived at its final position }
until j = k
:
//O(logn2 )
h, heap n , heap
h
h+1
20 + 21 +....+2h1 < n 20 + 21 +....+2h 2211 < n 2 211
n 2h+1 -1 = n+1 2h+1 = lg(n + 1) h + 1
h lg(n + 1) 1 = h=dlg(n + 1) 1e=blg nc
n>2h -1 n+1 > 2h h < lg n + 1 h-1 < lg(n + 1)-1 h

.
www.Bshams.ir
O(logn2 )
www.Bshams.ir
MAX HEAP
61
.3.3
procedure percolate(T[1..n],i)
{we suppose that T would be a heap if T[i] were sufficiently small,
we also suppose that 1i n the parametr n is not used here}
ki
repeat
jk
if ((j>1) && ( T[ 2j ] < T[k]) )
k 2j
exchange T[j] and T[k]
until j=k;
//O(log n)
........................................................
f unction f ind max( T[1..n] )
{ returns the largest element of the heap T[1..n]}
return T[1];
//(1)
........................................................
procedure delete max(T[1..n])
Delete the root}
T[1]T[n]
sift-Down(T[1..n-1],1)
//O(log n)
. ,
.
sift-Down
procedure insert N ode(T[1..n],v)

T[n+1] v
percolate(T[1..n+1],n+1)
{O(lg n)}
heap
P rocedure slow M akeHeap(T[1..n])

{this procedure makes the array T[1..n] into a heap}
for i 2 to n do
percolate(T[1..n],i)
www.Bshams.ir
// O(n log n)
www.Bshams.ir
.3
62
P rocedure M akeHeap(T[1..n])
For i b n2 c DownTo 1 Do
SiftDown(T, i)
Heap
O(n)
//O(n)
:
.
:
r+1 r
SiftDown
(.

Repeat
r)
n
K = blog2 c
Heap
:
.For
SiftDown
Repeat
:t
t 2 2k1 + 3 2k2 + . . . + (k + 1) 20
t 2k + 2k + 2 2k1 + 3 2k2 + . . . + (k + 1) 20
t < 2k + 2k+1 (21 + 2 22 + 3 23 + . . .)
P
t < 2k + 2k+1
n( 12 )n < 2k + 2k+1 2 t < 2k (22 1)
n=1
t < 3 2k t < 3n t O(n)
K = blogn2 c 2k ' n
1
2
1x = 1 + x + x + . . .
x
(1x)
2
n=1
1
2
(1x)
2 = 1 + 2x + 3x + . . .
P
x
= x + 2x2 + 3x3 + . . .
(1x)
nxn
2 =
n=1
n( 12 )n =
1
2
(1 12 )2
=2
( ) Binomial Heap
.
4.3
Binomial Heap
n2 n1 . heap heap : )
(. O(n1 + n2 ) heap .
www.Bshams.ir
www.Bshams.ir
( ) BINOMIAL HEAP .4.3
63
( )Binomial Tree
Bn1
1.4.3
.
.
Bn1
1
B1
B0
Bn =
(x + y)n
B2
B3
n=0
Bn1 Bn1
; ,
2n Bn
2n1 Bn
2n1 Bn
) .
kn
Bn
. ,
n Bn
Max Binomial Tree
2.4.3

b ,
.
www.Bshams.ir
Max Binamial Tree
www.Bshams.ir
.3
64
The Merge Of Max Binomial Trees

.
Bn+1 ,Max Binomial Tree Bn0
3.4.3
Max Binomial Tree Bn
merge
.
.
min
(1) merge
8
7
6
3
2
1
+
3
.5
9
7
6
4
4
3
8
2
1
6
3
2
.5
Binomial Heap
.
www.Bshams.ir
Binomial Tree
4.4.3
H ,Binomial Heap
www.Bshams.ir
( ) BINOMIAL HEAP .4.3
65
Binomial-Heap
Head[H]
- 10
- 1
- 6
12
25
8
11
18
14
17
29
38
27
Min Binomial Heap
Heap
1 .

,n
Binomial Tree
Binomial Heap
Binomial Tree
Node
Binomial Tree ,blg nc + 1
Min Binomial Heap
.2
.
.
.1
Heap
Node
5.4.3
Make-Heap()
x , Node
Node
Insert(H, x)
Minimum(H)
.
Delete
Heap
Node :
Extract-Min(H)
Node
Heap Heap
. Heap
www.Bshams.ir
www.Bshams.ir
Node
.3
66
Heap :
.
x ,Node
(x > k)
Union(H1 , H2 )
H2
H1 ,Heap
Decrease-Key(H, x, k)
x ,Node
Delete(H, x)
Max Binomial Heap

;
6.4.3
Max Binomial Heap

: .

4 ........ 9 ..................... 7
3
4
5
3
2
Max Binomial Heap
merge
. ,
www.Bshams.ir
(1)
www.Bshams.ir
FIBONACCI HEAP .5.3
67
R
8 ......... 6
4
?
4 .......... 9 ............. 7
3
4
2
4
5
?
8 ......... 9 ............. 7
4 3
5
1
6
5 2
4
4
3
1
2
1
Binomial Heap :
FIBONACCI HEAP
Fibonacci Tree
5.3
1.5.3
: :
Fn =
Fn2 , Fn1
www.Bshams.ir
n=0
n=1
else
www.Bshams.ir
.3
Fn2 ,Fn1
F2:
F1:
F0:
68
F3 :
F4 :
2.5.3
Max Fibonacci Tree

.
:
f0 = 0 , f1 = 1 , f2 = 1 , f3 = 2, ...
Fn fn+2
.
(n 0) fn
Fn
Fn
3.5.3
Fibonacci Heap

fn+1
Fibonacci Tree Fibonacci Heap
4.5.3
Max Fibonacci Heap

Max Binomial Heap

Fi+2
Fi+1
Max Fibonacci Tree
Fi+1
Fi
Fi
www.Bshams.ir
Merge
Merge
www.Bshams.ir
2-3 .6.3
69
Fi+1
Fi
Fi
Fi+1
Fi
SiftDown
.

.
lg
SiftDown
Height(Fn ) = d n2 e
. :
2-3
6.3
:
. 2-3
2-node
3-node
2-node
2-node
dataL.key
dataL.key
MiddleChild
MiddleChil
LeftChild
dataL.key
dataL.key
dataL.key
dataR.key
dataL,dataR
dataL.key < dataR.key
LeftChild
2-3
MiddleChild
2-3
dataR.key RightChild 2-3
FUNDAMENTALS OF DATA STRUCTURES IN C++,

Ellis Horowitz,Sartaj Sahni,Dinesh Mehta
www.Bshams.ir
LeftChild ,MiddleChild,RightChild
dataL
2-3 ,
LeftChild
3-node
2-3
3-node
www.Bshams.ir
.3
70
. 1 2-3
A
40
B
10 20
80
2-3 1
.
.
1 2h 1 n h 2-3
, .
,
, . ,
3-node 2-node
3-node
2-node
2-3 .
.
h=1
h = 2 .. X
...
.
.
..
h=h
20
X
X ..
...
...
...
21
2h1
2-node
m
l
h
n = 20 + 21 + + 2h1 n = 2211 = 2h 1 h = log2n+1
www.Bshams.ir
www.Bshams.ir
2-3 .6.3
71
h=1
h=2
X X
X .X
..
..
..
..
.
X .X
..
...
.
.
.
...
X .. X
...
...
...
..
h=h
2 30
2 31
2 3h1
3-node
h
n = 2 30 + 2 31 + + 2 3h1 n = 2(3311) = 3h 1 h =
l
m
1
logn+
3
l
m
l
m
1 < h < logn+1
logn+
3
2
: 2-3
templete<class KeyT ype >class Two3;//forward declaration

templete< class KeyT ype >
class Two3Node{
friend class Two3 < KeyT ype >;
private :
Element< KeyT ype > dataL,dataR;
Two3Node *LeftChild, *MiddleChild, *RightChild;};
class Two3{
public:
Two3(KeyT ype max,Two3Node< KeyT ype > *init=0)
:MAXKEY(max), root(init){};//constractor
Boolean Insert(const Element< KeyT ype > &);
www.Bshams.ir
www.Bshams.ir
.3
72
Boolean Delete (const Element< KeyT ype > &);

Tow3Node< KeyT ype > *Search(const Element < KeyT ype >&);
private:
Tow3Node< KeyT ype > *root;
KeyT ype MAXKEY;
};
M AXKEY
dataR.key = M AXKEY
MiddleChild
LeftChild
2node
RightChild
dataL
2-3
1.6.3
x Tow3::Search
3,2,1
compare
, ,
. ,
. 2-3
.
O(log n)
Tow3::Search
Tow3Node< KeyT ype > *Tow3< KeyT ype >::

Search(const Element< KeyT ype >& x)
//If the element x is not in the tree,then return 0,Otherwise
//return a pointer to the node that contains this element.
{
for(Tow3Node< KeyT ype > *p=root;p;)
for
www.Bshams.ir
www.Bshams.ir
2-3 .6.3
73
switch(p compare(x)){
case 1 :p=p LeftChild;break;
case 2 :p=p MiddleChild;break;
case 3 :p=pRightChild;break;
case 4 :return p ;// x is one of the keys in p
}
}
2-3
2.6.3
70 . f 2-3
. 1 2-3
f .

2-3
2-3 70 .
, 70 .
.
, 2-3
70 .
, . 80
() 2 .
.
. 30

3-node B
B A
x B
. () 2 .
www.Bshams.ir
www.Bshams.ir
.3
74
A
40
A
20 40
B
C
70 80
B
10 20
C
70 80
D
30
10
30 ()
70 ()
1 2-3 -2
2-3 60

60 . () 2
x ,
3-node
.(80)
(70) .(60)
3-node A
f .
. 20, 40, 70

40 . 2-3
2-3 ,
. 3
G
40
A
20
B
10
D
30
F
70
C
60
E
80
() 2 2-3 60 -3
www.Bshams.ir
www.Bshams.ir
2-3 .6.3
75
3-node
. .
.
q,p

Boolean Tow3< KeyT ype >::Insert(const Element< KeyT ype >& y)
//Insert the element y into the 2-3 tree only if it does not already
// contain an element with the same key.
{
Tow3Node< KeyT ype > *p;
Element < KeyT ype > x=y;
if(x.key>=MAXKEY) return FALSE; //invalid key
if(!root){NewRoot(x,0); return TRUE;} //empty 2-3 tree
if(!(p=FindNode(x))){
insertionError();
return FALSE;}//key already in 2-3 tree
for(Tow3Node< KeyT ype > *a=0; ; )
if(pdataR.key==MAXKEY){// p is a 2-node
pPutIn(x,a);
return TRUE; }
else{// p is a 3-node
Tow3Node< KeyT ype > *olda=a;
a=new (Tow3Node< KeyT ype >);
x=Split(p,x,olda,a);
if(root==p){ //root has been split
NewRoot(x,a);
return TRUE; }
else p=p parent();
}//end of p is a 3-node and for loop
}//end of Insert
www.Bshams.ir
www.Bshams.ir
.3
76
.
: ,
(
< KeyT ype >
: void Tow3::NewRoot(const Element& x,Tow3Node *a)

. 2-3
(new root)
( )
LeftChild
Middle Child, a
.
.
:Tow3Node* Tow3::Find(const Element & x)

. (1 )
.
Tow3::Search
x.key
2-3
.
.
FindNode
FindNode
. ( )
.
p root
root
Tow3::Insert
. ,

,
.
:void InsertionError()

. .
: void Tow3Node::PutIn(const Element& x,Tow3Node *a)

, f

(this)
Middle Child
RightChild
dataR
RightChild
dataL
MiddleChild
dataL
a dataR
:Eelement& Tow3::Split(Tow3Node* p,Element& x,Tow3Node *olda,*a)

www.Bshams.ir
(p)
Tow3Node
www.Bshams.ir
2-3 .6.3
77
ap
p ,
olda p
. .
for
x , Insert
2-3 . 2-3
.
O(log n)
2-3
3.6.3
. 2-3

.
() 4 50
60 20
.

. ,
.
70 . () 4
.
dataR.key
MAXKEY , C
2-3 90 . () 4
dataL
dataR
(dataR.key=MAXKEY)
() 4
MAXKEY
dataR.key
. () 4 2-3
www.Bshams.ir
www.Bshams.ir
.3
78
A
50
C
B
10
80
D
60
20
70
90
95
2-3 ()
A
50
B
10 20
60
80
D
90 95
70 ( )
A
50 80
B
10
C
60
20
95
90 ()
, 4
. 60
3-node B
50 ,
dataL
20
dataR.key=MAXKEY
. , () 4 2-3
www.Bshams.ir
www.Bshams.ir
2-3 .6.3
79
(rotation)
60 .
2-node C
, 95
rotation
80 .
(combine)
95 .
50 . () 4 2-3
. () 4 2-3
10
2-node C
2-node
3-node C
.3-node
, 60
2-node
80 20 . 95
b .
(60 ) ,
,
. (95 )
.(() 4 )
www.Bshams.ir
www.Bshams.ir
.3
80
A
20 80
B
10
C
95
50
60 ()
A
20
B
10
C
50 80
95 ()
A
20
C
B
10
80
50 ()
B
20 80
10 ()
4
2-3

:1
.
:2
for(;p has zero element && p!=root;p=r){

left r be the parent of p and let q be the left or right sibling of p ;
if(q is a 3-node) perform a rotation
else perform a combine;}
www.Bshams.ir
www.Bshams.ir
2-3 .6.3
81
,
.
q , r
:3
.
.
q,
3-node 2-node
? . 5
)
c,b,a.

.(
r
X ?
q
Y Z
p
X
a
d
r
r
Y ?
q
Z
c
()
r
Y ?
r
Z ?
q
X Y
a
c d

a
r p
r
W Z
a q
b
r
a q
X
d e

b
()
r
W Y
X Y
b
p
Z
q
X
Z
c
p
()
2-3 -5
www.Bshams.ir
www.Bshams.ir
.3
82
combine
r
x
p
x y
a
( )
r
x z
q
y
q
y
c
r
z
p
x y
a
d
c
( )
2-3
:
template < class KeyT ype >

Two3< KeyT ype >::DeleteKey(Two3Node< KeyT ype > *p,
const Element< KeyT ype >& x)
//Key x.Key is to be deleted from the leaf node p.
{
if( x.Key==pdataL.Key) // first element
if(pdataR.Key!=MAXKEY )
{ // p is a 3-node
p dataL = p dataR;
p dataR.Key =MAXKEY;}
else p dataL.Key=MAXKEY; // p is a 2-node
else p dataR.Key =MAXKEY;} // delete second element
www.Bshams.ir
www.Bshams.ir
2-3 .6.3
83
Combine Rotation
//Rotation when p is the left child of r and q is the middle child of r.

p dataL= r dataL;
p MiddleChild = q LeftChild;
r dataL = q dataL;
q dataL = q dataR;
q LeftChild = q MiddleChild;
q MiddleChild = q RightChild;
q dataR.Key = MAXKEY;
...............................................................
//Combine when p is the left child of r and q is the right sibling of p.
p dataL= r dataL;
p dataR= q dataL;
p MiddleChild = q LeftChild;
q RightChild = q MiddleChild;
if(r dataR.Key = MAXKEY)// r was a 2-node
r dataL.Key = MAXKEY;
else {
r dataL= r dataR;
r dataR.Key = MAXKEY ;
r MiddleChild = r RightChild;
}
2-3
O(1)
4.6.3
. , .
. 2-3
2-3

2-3
.
www.Bshams.ir
O(log n)
www.Bshams.ir
.3
84
Red-Black -
3 :
7.3
1.7.3
nil
, ) . ,
( f )
( ).
26
17
41
14
10
7
3
NIL
16
15
12
19
NIL NIL
23
15
47
30
NIL
28
NIL
NIL
38
NIL
35
NIL
39
NIL
NIL
NIL
NIL
NIL NIL
NIL
NIL
www.Bshams.ir
NIL
NIL NIL
NIL
NIL
21
www.Bshams.ir
RED-BLACK - .7.3
85
bh(X) Black-Height(x)
4
.
2.7.3
:Black-Height(x)
: uncle(x)
if parent[x]=right[parent[parent[x]]] then
uncle[x]:=left[parent[parent[x]]]
else uncle[x]:=right[parent[parent[x]]]
n
RB
: 1
1)
2 log(n+
2
:
.
2bh(x) 1 x
:
.
nil
20 1 = 0 2bh(x) 1 x
:
x bh(x)
Black - Height,
bh(x)-1
bh(x)
2bh(x)1 1 x
bh(x)1
. 2
1 + 2bh(x)1 1 + 1 = 2bh(x) 1
b .
. , (
Black-Height .
3
.
(n+1)
n 2 1 = 2 n + 1 = h 2 log2
h
O(log n) ,RB
. b x
www.Bshams.ir
www.Bshams.ir
.3
86
3.7.3
RB

( ) .
.
.
:
.

O(1)
nil
Right-Rotate(T,y)
-
y
x
Left-Rotate(T,x)

Left-Rotate(T,x)
1. y right[x]
// SET y
2. right[x] left[y]
// Turn ys left subtree into xs right subtree
3. if left[y] 6= NIL then

4.
p[left[y]] x
5. p[y] p[x]
// Link xs parent to y
6. if p[x] =Nil then

7.
root[T] y
8.
else if x= left[p[x]]then
9.
left[p[x]] y
10.
else right[p[x]] y
11. left[y] x
12. p[x] y
www.Bshams.ir
// Put x on ys left
www.Bshams.ir
RED-BLACK - .7.3
87
y=uncle[x]
4.7.3
5 bh
4
3
RB
.
.
6
:
Parent[Parent[x]]
RB-insert
D
Bx
new x C
Dy
x A
new x C
Parent[x]
Black-Height
y=uncle[x]
www.Bshams.ir
5
6
www.Bshams.ir
.3
A
Bx
x A
88
?
B
x A
RB-Insert(T,x)
1. Tree-Insert(T,x)
2. color[x] RED
3. while x6=root[T] and color[p[x]]=RED do
4.
if p[x] =left[p[p[x]]] then
5.
y right[p[p[x]]]
6.
if color[y] =RED then
7.
color[p[x]] BLACK
// Case 1
8.
color[y] BLACK
// Case 1
9.
color[p[p[x]]] RED
// Case 1
10.
x p[p[x]]
// Case 1
11.
else if x = right[p[x]] then
12.
x p[x]
// Case 2
13.
LEFT-Rotate(T,x)
// Case 2
14.
color[p[x]] BLACK
// Case 3
15.
color[p[x]] RED
// Case 3
16.
RIGHT-Rotate(T,p[p[x]])
// Case 3
17.
else(same as then clause with right and left exchanged)
18. color[root[t]]BLACK
www.Bshams.ir
www.Bshams.ir
RED-BLACK - .7.3
89
11
11
14
15
case 1-
2
1
y
15
x 4
14
8
4
case 2
?
11
x
2
1
case 3
11
5
x 2
14
1
4
14
15
y
15
5
4
RB-Delete
5.7.3
nil[T]
.
.

Parent ) x

(.
while
nil
nil[T]
nil
RB-Delete-Fixup
RB
Parent
(.
,
.

www.Bshams.ir
nil[T]
nil
RB
RB
www.Bshams.ir
.3
90
. 7
.
nil[T].Parent,
16 ,
RB
nil
Parent[x]
RB-Delete-Fixup
17
.
RB-Delete
. (Successor)
.
.
Key[z]:=Key[Succ(z)]
,
. 4
(
. ,
) .
. 1 (

:
.
. ,
. 1 (
4

: (. 8
.
w=sibling(x)
,
.
www.Bshams.ir
c,c
www.Bshams.ir
RED-BLACK - .7.3
91
( ) w(1
. w(2
. w(3
. w(4
....
..
B
....
..
D
case 1
x A
(a)
case 2
x A
www.Bshams.ir
case 4
A
....
.. c
D
w
D
...
...
c
B
(d)
new w
C
xA
....
.. c
B
case 3D
...
...
Bc
new x C
D
C
(c)
....
.
new x . c
B
...
...
B c
(b)
c
C

new x =root[T]
www.Bshams.ir
.3
92
1 : (1
. 4 3 2
: (2
.
.
3 : (3
. 4
4 (4
O(lg n)
RB-Delete(T,z)
1. if left[z]=nil[T] or right[z]=nil[T] then
2.
yz
3.
else y Tree-Successor(z)
4. if left[y] 6= nil[T] then

5.
x left[y]
6.
else x right[y]
7. p[x] p[y]
8. if p[y] = nil[t] then
9.
root[T] x
10.
else if y = left [p[y]] then
11.
left[p[y]] x
12.
else right[p[y]] x
13. if y 6= z then
14.
15.
key[z] key[y]
// if y has other fields , copy them , too
16. if color[y] = BLACK then
www.Bshams.ir
www.Bshams.ir
RED-BLACK - .7.3
93
17.
RB-Delete-Fixup(T,x)
18. return y
RB-Delete-Fixup(T,x)
1. while x 6= root[T] and color[x] =BLACK do
2.
if( x = left[p[x]]) then
3.
w right[p[x]]
4.
if( color[w]=RED) then
5.
color[w] BLACK
//case1
6.
color[p[x]] RED
//case1
7.
LEFT-Rotate(T,p[x])
//case1
8.
w right[p[x]]
//case1
9.
if(color[left[w]]=BLACk and color[right[w]]=BLACK) then//case1
10.
color[w]=RED
//case2
11.
x p[x]
//case2
12.
else if color[right[w]]=BLACK then
13.
color[left[w]] = BLACK
//case3
14.
color[w] RED
//case3
15.
RIGHT-Rotate(T,w)
//case3
16.
wright[p[x]]
//case3
17.
color[w]color[p[x]]
//case4
18.
color[p[x]]BLACK
//case4
19.
color[right[w]] BLACK
//case4
20.
LEFT-Rotate(T,p[x])
//case4
21.
xroot[T]
//case4
22.
else(same as then clause with right and left exchanged)
23. color[x] BLACK
www.Bshams.ir
www.Bshams.ir
.3
94
(Disjoin sets)
.
...21
AB
8.3
n...21
.

.
set
set={1 , 2 , 3 , 4 , 5 , 6 , 7 , 8 , 9 , 10 }
set
1
10
10
1={1 , 5 }
2={2 , 4 , 7 , 10 }
3={3 , 6 , 8 , 9 }
3
1
4
set
10
set
1
10
function find1(x)
{ find the lable of the set containing x}
return set[x]
x find1
.
b a merge1
.
www.Bshams.ir
(1)
(n)
www.Bshams.ir
(DISJOIN SETS) .8.3
95
function merge1(a,b)
{merge the sets labled a and b }
imin(a,b)
jmax(a,b)
fork1 to n do
if set[k] = j then set[k] i
merge2 find2
function find2(x)
r x
while(set[r]6= r) do
rset[r]
return r
if(a<b) then set[b] a
else set[a] b
merge2
O(n) find2
(1)
find
.

.
:
if(height[a] = height[b])
www.Bshams.ir
find3
merge3
www.Bshams.ir
.3
96
set[b] a
height[a] ++
r a
else
if(height[a]>height[b]) then
set[b] a
ra
else
set[a] b
rb
return r
function find3(x)
rx
while(set[r]6=r) do
r set[r]
ix
while i6= r do
j set[i]
set[i] r
ij
return r
.
www.Bshams.ir
log n
www.Bshams.ir
.4

:
www.Bshams.ir
www.Bshams.ir
.4
98
(Greedy Algorithms)
1.4
.

.
.

.
:
Control Abstraction
Fuction Greedy(C):set //solution set

{ C is a set of condidates}
S { S C is a set of solution ,if it is possible }
{ Greedy loop}
While (C 6= & not Solutions(S))
x Select(C)
CC\{x}
if feasible ( S { x }) then
SS{x}
if (solution(S)) then return S
else
there are no solutions.
solution , feasible
Select
feasible
Select
b .
1.1.4
.

:
www.Bshams.ir
www.Bshams.ir
(GREEDY ALGORITHMS) .1.4
99
,

.

.
,
.
.
,
.
7000
1100,
1800 ,C 900,
1200,
1800
200
7000
1
4000
0
0
2200
0
3000
0
www.Bshams.ir
1
1
1100
1
1200
1800
900
1100
1800
200
A=00
B=01
C=1000
E=101
D=11
M=1001
www.Bshams.ir
.4
100
. ,
. ,
, 8
4
: .
2 1200 +2 1800+4900+31100+21800+4200=17300
.
70008=56000
DBAD
11010011
01001001101
,

.
BAME

.

.
n-1
www.Bshams.ir
. i-1 1
...
.. .
..
15 ...
....
....
....
....
....
7
....
....
....
....
....
....
....
....
2
4
8
www.Bshams.ir
101
MST

2.1.4
G=<N,A>
.A

.( )
:
.(
T=<V,E>
p) p=q+1 , T
.1
. p=q+1 T
.2
. .3
. .4
pp2
. (
p2
Kp
: )?
(.
| {z } |
K3
{z
K4
332 442 = 48

(minimum spaning tree)
www.Bshams.ir
www.Bshams.ir
.4
102
f :

.
:
Prim , Kruskal, Boruvka ,Sollin
:kruskal
:

find
n-1
greedy ( n 1
find(v) ,find (u)
e=uv
.

.
find(v),find(u)
n-1 T
function kruskal(G=< N,A > : graph,length :A R+ ): set of edge

sort A by increasing length
initialize n sets , each set containing an element of N
T
{greedy loop}
repeat
e{u,v}
A A \ {e}
ucompfind (u)
vcompfind (v)
www.Bshams.ir
www.Bshams.ir
103
if(vcomp 6= ucomp) then

TT {e}
merge(ucomp,vcomp)
until T is containing n-1 elements
return T
1
6
4
3
6
8
4
3
6
4
7
step
initialization
connected components
{1}{2}{3}{4}{5}{6}{7}
{1,2}
{1,2}{3}{4}{5}{6}{7}
{2,3}
{1,2,3}{4}{5}{6}{7}
{4,5}
{1,2,3}{4,5}{6}{7}
{6,7}
{1,2,3}{4,5}{6,7}
{1,4}
{1,2,3,4,5}{6,7}
{2,5}
{4,7}
.n log n
reject
{1,2,3,4,5,6,7}
n , log n merge,find
: .
(n 1) a
|A| = a
n(n1)
log (n 1) log a log n + log (n 1) log 2

log n log a 2 log n
.
www.Bshams.ir
O(a log a) sort
log n
log a
www.Bshams.ir
.4
104
: Prim
:

G B
b , ,
( )

binary heap
fibonacci heap
O(n2 )
O((a + n)logn)
O(a + nlogn)
.
function Prim(G=< N, A >:graph,length:A R+ ):set of edges

B {an arbitary element of N}
T
while B 6= N do
e { u,v} such that e is minimum and u B,v N\B
T {e} T
B B {v}
:
function Prim(L[1..n,1..n]):set of edges

1. {initialization : only node 1 is in B}
2.T {will contain the edges of the minimum spanning tree}
3. for i=2 to n do
4.
nearest[i] 1
5.
mindist[i] L[i,1]
6. {greedy loop}
7. repeat n-1 times
8.
min
9.
for j 2 to n do
10.
www.Bshams.ir
if 0 mindist[j] < min Then
www.Bshams.ir
105
11.
min mindist[j]
12.
kj
13.
T=T {nearest[k],k}
14.
mindist[k] -1{add k to B}
15.
for j 2 to n do
16.
if L[j,k] < mindist[j] Then
17.
mindist[j] L[j,k]
18.
nearest[j] k Return T
: heap PRIM
HEAP-PRIM(G)
1. A
2. for each x V do key[x]
3. H Build-heap(key)
4. select arbitrary vertex v
5. S {v}
6. Decrease-key(H,v,0)
7. Extract-min(H)
8. for each x adjacent to v do
decrease-key(H,x,w(v,x)), [x] v
9. for i=1 to |V| - 1 do
10.
begin
11.
Extract-min(H),let v be the vertex
12.
A A {(v,[v])}
13.
S S {v}
14.
for each x V-S adjacent to v do
15.
if w(x,v) < key[x] Then
16.
17.
Decrease-key(H,x,w(x,v)),[x] v
end
18. return A
www.Bshams.ir
www.Bshams.ir
.4
106
heap
:binary-heap
key[x]
heap
heap
Extract-Min
Build-heap
O(|V |)
heap
, ,
(1) heap
heap
O(log|V |)
heap
heap
Decrease-Key
O(|E|log|V | + |V |log|V |) =
heap
O(log|V |)
O((|E| + |V |)log|V |)
:fibonacci-heap(amortized)
Extract-Min ,

O(1)
O(|V |)
Decrease-Key
Build-heap
O(log|V |)
O(|E| + |V |log|V |)
|E| = (|V |)
fibonacci-heap
|E| = (|V |2 )
fibonacci-
(dense graph)
|V |
|E|
O(|V |log|V |)
O(|V |2 log|V |) binary-heap
.
.
O(|V |2 ) heap
fibonacci-heap
www.Bshams.ir
O(|V |2 )
www.Bshams.ir
107
: Sollin
b ,

4| , {z
3, 2} 1|{z}
, 6 ,5, 0
|{z}
C
C B
241816
16 2528
. 25 2825
28
0
10
24
14
10
2
16
2
18
25
4
0
16
14
12
22
12
25
4
22
algorithm Sollin(G)
begin
Fo =(v,k);
i=0;
while there is more than one tree Fj do
for each tree Tj in forest Fj do
choose the minimum weighted edge (u,v) joining some
vertix u in Tj to a vertix v in some other tree Tk in forest Fj
from the otherforest Fj+i by joining all Tj and Tk of Fj with tree
converesponding select edge
i++;
end;
www.Bshams.ir
www.Bshams.ir
.4
108
: Boruvka
Boruvka( G = (V, E) , w ) {
initialize each vertex to be its own component ;
A = {} ; //A holds edges of the MST
do {
for ( each component C ) {
find the lightest edge (u,v) with u in C and v not in C ;
add { u, v } to A ( unless it is already there ) ;}
apply DFS to graph H = (V, A) , to compute the new components
} while (there are 2 or more components) ;
return A ;} // return final MST edges
BORUVKA(V,E):
F=(V,)
while F has more than one component
choose leader using DFS
FIND-SAFE-EDGES(V , E)
for each leader v add safe ( v ) to F
FIND-SAFE-EDGES(V, E):
for each leader v
safe(
v )
for each edge (u,v) E
u
leader(u)
v leader(v)
if u
6= v
if (w(u,v) < w( safe( u
))
safe ( u
) (u, v)
if (w (u,v) < w(safe(
v ))
safe ( v) (u, v)
www.Bshams.ir
www.Bshams.ir
109
: Knapsack
.
,
3.1.4
1, 2, ..., n
vi
wi

: .
xi wi W
, xi [0, 1]
P
max [
xi vi ]
vi
wi
weight
x[k] =
W weight

W [k]
O(n log n)
max
function Knapsack(w[1..n],v[1..n],W):array[1..n]
for i 1 to n do
x[i] 0
weight 0
{gready loop}
while (weight<W) do
i the best remaining object
if(weight+w[i] W )then
x[i]1
weightweight+w[i]
else
x[i]
(W weight)
w[i]
weight W
return x
www.Bshams.ir
www.Bshams.ir
.4
110
: DIJKSTRA
4.1.4
.
.
. b b
.
function Dijkstra(L[1..n,1..n]):array D[2..n]

C{2,..,n} {s=N\C exists only implicity}
for i 2 to n
D[i] L[1,i]
repeat n-2 times
v some element of C minimizing D[v]
C C\{v}
for each w C do
D[w] min(D[w], D[v]+L[v,w])
return D
weight of edge
0
Lij =
(i,j) E
i=j
else
:
1
10
50
=
5
10
30
100
U
4
20
W
-3
R
32

5
50
www.Bshams.ir
step
initilization
2,3,4,5
[50,30,100,10]
2,3,4
[50,30,20,10]
2,3
[40,30,20,10]
[35,30,20,10]
www.Bshams.ir
111
.

:
function Dijkstra(L[1..n,1..n]):array D[2..n] , array P[2..n]

C{2,..,n}{s=N\C exists only implicity}
for i 2 to n
D[i] L[1,i]
P[i]0
repeat n-2 times
v some element of C minimizing D[v]
C C\{v}
for each w C do
if ( D[w]>D[v]+L[v,w] ) then
D[w] D[v]+L[v,w]
P[w] v
return D , P
:
:
(n 2) + (n 3) + . . . + 2 + 1 =
(n 1)(n 2)
(n2 )
:
: .
2
P 3
3 4
0 5
5
0
b 1 b
f , .
4 b
5 4 5 4 ,
. 4 5 1 4 5

O(a+nlogn)
Fibonacci-heap
www.Bshams.ir
O((a+n)logn) Binary-heap
www.Bshams.ir
.4
112
(timetable or scheduling)
5.1.4
server
minimum
.

.

(with deadline)
(simple)
server
. .

ti (1 i n)
server

.
.
Ti = t1 + t2 + ... + ti1 + ti : i
P Ti
T =
n :
P
T T
.
:
, 5 . 3
server
. 6 3 3 10
1 , 2, 3
1, 3, 2
2, 1, 3
2, 3, 1
3, 1, 2
3, 2, 1
www.Bshams.ir
= T = 5 + (5 + 10) + (5 + 10 + 3)
= T = 5 + (5 + 3) + (5 + 3 + 10)
= T = 10 + (10 + 5) + (10 + 5 + 3)
= T = 10 + (10 + 3) + (10 + 3 + 5)
= T = 3 + (3 + 5) + (3 + 5 + 10)
= T = 3 + (3 + 10) + (3 + 10 + 5)
www.Bshams.ir
113

.
.
O(n log n)
(Scheduling with deedline)

di
(T=1,2,...)
gi 0
.
.
:

deedline
i
1
2
3
4
gi 50 10 15 30
di
2
1
2
1
sequence profit sequence
1
50
2,1
2
10
2,3
3,1
3
15
4
30
4,1
1,3
65
4,3
profit
60
25
65
80 optimum
45
fisible
fisible , 2,1
fisible
fisible
.

,
fisible , 1,2
1, 2, ..., k
1, ..., k fisible , j . d1 d2 ... dk

.
www.Bshams.ir
fisible
fisible
www.Bshams.ir
.4
114
deedline
Function sequence(d[0..n]):k , array[1..k]

array j[0..n]
{The schedule is constructed step by step in the array j ,the variable k
say how many jobs are already in the schedule}
d[0]j[0]0 {sentinels}
k j[1]1{job 1 is always choosen}
{gready loop}
for i 2 to n do {decreasing order of g}
rk
while d[j[r]]>MAX(d[i],r) do r r-1
if d[i]> r then
for m k step -1 to r+1 do
j[m+1]j[m]
j[r+1]i
k k+1
return k,j[1..k]
:
i
gi
di
1
20
3
2
15
1
3
10
1
3
Initialization:
1 3
Try 2:
Try 3: unchanged
1 3 3
www.Bshams.ir
4
7
3
5
5
1
6
3
3
www.Bshams.ir
115
Try 4:
Try 5: unchanged
Try 6: unchanged
= optimal sequence: 2 , 1 , 4
value=42
.
O(n2 )
: Disjoin Set
b
f ,
merge find

.
Function sequence2(d[1..n]):k,array[1..k]
array j ,F[0..n]
{initialization}
for i 0 to n do
j[i] 0
F[i] i
initialize set [i]
{gready loop}
for i 1 to n do {decreasing order of g}
k find(min(n,d[i]))
m F[k]
if m 6= 0 then
j[m] i , l find(m-1)
F[k]F[l]
merge(k,l)
k0
for i 1 to n do
if j[i]>0 then kk+1 , j[k]j[i]
return k,j[1..k]
www.Bshams.ir
www.Bshams.ir
.4
116
.
:
Initilize : L=min(6, max(d,1))= 3

F=
0
0
1
1
2
2
Try 1: d1 =3 , assign task 1 to position 3

F=
0
0
1
1
2
2
M
3
Try 2: d2 =1 assign task 2 to position 1

F=
0
0
2
K
K
3
Try 3: d3 =1 no free position available since the F value is 0

Try 4: d4 =3 assign task 4 to position 2
F=

0
0
o
2
]
3
Try 5: d5 =1 no free position available

Try 6: d6 =3 no free position available
= optimal sequence: 2 , 1 , 4
value=42
.
www.Bshams.ir
O(nlogn)
www.Bshams.ir
(DEVIDE AND CONQUER) .2.4
117
(devide and conquer)
2.4
D&C
b f b
b . b

. b
:
D& C
function DC(x)
if x is sufficiently small or simple then return adhoc(x);
decompose x into smaller instances x1 , x2 , ..., xl ;
for i1 to l do yi DC(xi );
recombine the yi s to obtain a solution y for x;
return y;
1.2.4
? , .
O(n2 )
981 1234 = (|{z}

9 102 + |{z}
81 ) (|{z}
12 102 + |{z}
34 ) =
w
(w 102 + x)(y 102 + z) = wy 104 + (wz + xy) 102 + xz

: .
T (n) 4T ( 2 ) + (n) T (n) O(n2 )

2
O(n )
D&C
(wz + xy) = (w + x)(y + z) wy |{z}

xz = r p q
{z
} |{z}
|
r
wy 104 + (wz + xy) 102 + xz = p 104 + (r p q) 102 + q
www.Bshams.ir
www.Bshams.ir
.4
118
T (n) 3T ( n2 ) + (n) = T (n) O(nlog2 )

2 +1 2
(n)
large-integer prod(large-integer u , large-integer v)

{
large-integer x,y,w,z,r,p,q
int m , n
n=Maximum(number of digits in u , number of digits in v )
if(u==0 || v==0)
return 0
else if(n threshold )
return uv obtained in usual way
else
{
m=b n2 c
x=u divide 10m ;
y= u rem 10m
w=v divide 10m ;
z= v rem 10m
r=prod(x+y,w+z)
p=prod(x,w)
q=prod(y,z)
return p102m + (r-p-q)10m+q
}
}
www.Bshams.ir
www.Bshams.ir
119
merge sort
2.2.4
procedure merge-sort (T[1..n])

if n is sufficiently small then insertion sort(T[1..n])
else{
//arrayu[1 . . . b n2 c + 1], array
u[1 . . . b n2 c] T [1 . . . b n2 c]
v[1 . . . d n2 e + 1]
v[1 . . . d n2 e] T [1 + b n2 c . . . n]
merge sort(u[1 . . . b n2 c])
merge sort(v[1 . . . d n2 e])

merge(u,v,T)}
insertion sort
procedure merge(u[1..m+1],v[1..n+1],T[1..m+n]){
i,j 1
u[m+1]=v[n+1]
for k 1 to m+n Do
if u[i]<v[j]
T[k]u[i]
i++
else
T[k]v[j]
j++}
www.Bshams.ir
www.Bshams.ir
.4
120
:
:
1 2 3
6 7 8
9 10 11
31 3 10 54 7 17 12 43 48 8 2
31 3 10 54 7
31 3
3 31
3 31
10
3 31
17 12 43 48 8
10 54 7
17 12 43
10
17
7 54
7 54
48 8 2
12 43
2 8
48
1243
17
7 10 54
12 17 43
3 7 10 31 54
48 2 8
2 8
48
8 12 17 43 48
2 3 7 8 10 12 17 31 43 48 54
T (n) = T (d n2 e) + T (b n2 c) + (n)
T (n) = 2T ( n2 ) + (n)
= T (n) O(nlogn)
inplace
if u[i] < v[j]
stable
stable
if u[i] v[j]
www.Bshams.ir
www.Bshams.ir
121
Quick Sort
3.2.4
Procedure pivot(T[1. . .n] , var L)

{permutes the elements in array T[i . . .j] and return a value L such that
at the end,i L j, T [k] < p for all i k < L, T[L]=p and
T[k] > p for all L<kj , where p is the initial value of T[i] }
p T[i]
ki , Lj+1
repeat kk+1 until T[k] > p or k j
repeat LL-1 until T[L] p
while(k < L) do
swap(T[L],T[k])
repeat k k+1 until T[k] > p
repeat L L-1 until T[L] p
swap(T[i],T[L])
procedure quicksort (T[i. . .j])

{sorts subarray T[i. . .j] into nondecreasing order}
if (j - i) is sufficiently small then sort with insertion-sort(T)
else
pivot(T[i . . . j],L)
quicksort(T[i . . . L -1])
quicksort(T[L+1 . . . j])
17
48
10
p=T[i]=5
ki
L j+1
www.Bshams.ir
25
11
www.Bshams.ir
17
48
.4
48
48
122
25
17
25
17
25

L k
.
3
|
5
{z
48
} |
17
L L k
25
{z
}
:
pivot
T(n)=T(n-1)+(n)
T(n)=c
(n)
= T(n)=(n2 )
n n0
pivot
(n-i+1)
pivot

: ,
T(n)' 2T ( n2 ) + (n) T (n) = (n log n)
www.Bshams.ir
www.Bshams.ir
123
2 n+1 , quick sort
insertion sort
pivot
?
:
n+1+1
n+1
n+2
2
O((2 n + 1) ) O(n)
insertion
T(n-( n+2))
(n)
}|
{
z
}|
{
z
T(n) T ( n) + T (n n) + (n) + O(n)
k1
nk
z }| { z
}|
{
1 . . . k (k + 1) . . . n
T (n) = T (k 1) + T (n (k + 1) + 1) + (n) =
n
1 P (T (k
k=1
1) + T (n k)) + (n)
n
n
P
P
= 1n
T (k 1) + n1
T (n
k=1
k=1
n
n
n
P1
P1
P1
1
1
2
0
k) + (n) = n
T (k) + n
T (k ) + (n) = n
T (k) + (n) =
k=0
k=0
k=0
n1
n1
2
2 P T (k) + (n) = 2 a + 2 P T (k) + (n)
n (T (0) + T (1)) + n
n
n
k=2
k=2
www.Bshams.ir
www.Bshams.ir
.4
(n lg n)
124
n=2
. ,
(n lg n) = (2 lg 2) = 2
. 1
T (n) =
X1
2 a + 2 n
n
T (n) =
T (k) + (n) =
k=2
(k lg k)
X1
2 a + 2 n
n
k < n
(klogk) + (n)
k=2
X1
2 a + (( 2 ) n
(klogk)) + (n)
n
2 a + ( 2
n
k=2
n1
xlogxdx) + (n)
n 2
T (n) = (nlogn)
:

quick sort
.
:
.
quick sort
:
. 5

(dn/5e)
(bn/5c) n
5
.
n mod 5
5
.
insertion sort
.
,
select
pivot
www.Bshams.ir
ni
select
www.Bshams.ir
125
k i
3( 12 d n5 e 2) 310n 6
x : n ( 310n 6) = 710n + 6
x
5

.
3n 6
10
dn/5e
T (dn/5e)
7n + 6
10
O(1)
insertion sort
O(n)
T (n) = T (dn/5e) + T (7n/10 + 6) + O(n)

T (n) O(n)
2(n/5) 1
.
1 ((n 1) 2( n 1)) + 2( n 1) = 7n 3
2
5
5
10
2
= T (n) T (d n5 e) + T ( 710n 32 ) + O(n)
www.Bshams.ir
www.Bshams.ir
.4
126
( )
4.2.4
, 2
b
.
(n3 )
b
.
(nlog2 )

. 7 8
c11
c21
C=
c31
c41
c12
c22
c32
c42
c14

c24
=
c34
c44
c13
c23
c33
c43
A=
nn
a%
11
a21
a12
a22
c011
c021
c012
c022
B A, n n
22
B=
nn
b11
b21
b12
b22
,C = A B
nn
m1 = (a21 + a22 a11 )(b22 b12 + b11 )

m2 = a11 b11
m3 = a12 b21
m4 = (a11 a21 )(b22 b12 )
m5 = (a21 + a22 )(b12 b11 )
m6 = (a12 a21 + a11 a22 )b22
m7 = a22 (b11 + b22 b12 b21 )

m2 + m3
m1 + m2 + m4 m7
n
22
m1 + m2 + m5 + m6
m1 + m2 + m4 + m5
18 7

n 2
T (n) = 7T ( n2 ) + 18 2

18 n2 2 O(nlog27 ) = T (n) O(nlog27 )
.
www.Bshams.ir
www.Bshams.ir
127
:
. 24
?
T (24) = 7T (12) = 72 T (6) = 73 T (3) = 73 33

:
an
n 2
(a 2 )
an = a an1
n
n
n=1
: .
n
T ( 2 ) + 1
T (n) = T (n 1) + 1
T(n) :
n
n
n=1
n
T (b 2 c) + 1
T (n) = T (n 1) + 1 = T ( n1 ) + 2 = T (b n c) + 2
2
2
T (n) = T (b n2 c) + (1)
T (n) = (logn)
www.Bshams.ir
, (1) = (nlog2 1 (logn)0 )
n
n=1
www.Bshams.ir
.4
128
Dynamic Programming
3.4
b
b
b .
b
.

.
.
D&C
DP
. b
Bottom - Up
:
n
k
n1
k
n1
k1
k = 0 or
else

n
k
1.3.4
k=n
5
2
4
1
3
0
?
1
3
.1..
3
1..
4
2
3
2
.. ..
.. .
... ... .... ...
2
1
1
0
1
1
?
1
2
2
?
1
?
1
n
k
T(n,k)=
www.Bshams.ir
k = 0 or
T (n 1, k) + T (n 1, k 1) + 1
k=n
else
n
k

)
www.Bshams.ir
DYNAMIC PROGRAMMING
129
.3.4
g(n, k) = T (n, k) + 1:
g(n,k)=
1
k = 0 or k = n
g(n 1, k) + g(n 1, k 1)
else
g(n, k) = T (n 1, k) + T (n 1, k 1) + 1 + 1 = (T (n 1, k) + 1) +

(T (n 1, k 1) + 1) = g(n 1, k) + g(n 1, k 1) g(n, k) = nk

T (n, k) = nk 1
:
T(n,k)=
1l
k = 0 or k = n
T (n 1, k) + T (n 1, k 1) + l
else
n
k
T (n, k) =
1 l 0

:
:
c[n,k]=c[n-1,k-1]+c[n-1,k]
.
c
0
0
1
3
.
.
.
.
.
n-1
1
.
.
.
.
.
1
www.Bshams.ir
k-1
c[n-1,k-1]
(nk)
c[n-1,k]
c[n,k]
www.Bshams.ir
.4
dn ,. . . d2 ,d1
130
2.3.4
c[i,j]
di , . . . , d1
c[i,j]= 1+c[1 , j-d1 ]
c[i-1 , j]
min{ c[i-1 , j],1+c[i , j - di ]}
j=0
i=1 , j < d1
i=1 , j d1
i > 1 , j < di
i >1 , j di
: 8
amount
d1 =1
d2 =4
d3 =6
0
0
0
0
1
1
1
1
2
2
2
2
3
3
3
3
4
4
1
1
5
5
2
2
6
6
3
1
7
7
4
2
8
8
2
2
:
Function coins(N,n)
{array d[1..n] specifies the coin,in example there are 1,4,6 units}
array d[1..n]
array C[0..n,0..N]
for i 1 to n do
C[i,0]0
for i 1 to n do
for j 1 to N do
if(i=1 and j<d[1]) then C[i,j]
else if(i=1 and j d[1]) then C[i,j] 1+C[1,j-d[1]]
else if(i>1 and j<d[i])then C[i,j] C[i-1,j]
else C[i,j] min{C[i-1,j],1+C[i,j-d[i]]}
.
www.Bshams.ir
N+1 n
((N + 1)n)
www.Bshams.ir
DYNAMIC PROGRAMMING
131
{0, 1} b
.
vi
wi
.3.4
3.3.4
1,2,. . . n
.( )
.x i
{0, 1}

: .
max(
v[i,j]
xi wi W
1,2,. . . i
or 0
V[i,j]= V1
V[i-1,j]
max{ V[i-1,j],Vi +V[i-1,j-wi ]}
xi vi )
j=0, i
i=1 , 0<j < w1
i=1 , j w1
i > 1 ,0< j < wi
i >1 , j wi
weight
w1 =1
w2 =2
w3 =5
w4 =6
w5 =7
unit
v1 =1
v2 =6
v3 =6
v4 =22
v5 =28
0
0
0
0
0
0
1
1
1
1
1
1
2
1
6
6
6
6
3
1
7
7
7
7
((W + 1)n)
4
1
7
7
7
7
5
1
7
18
18
18
6
1
7
19
22
22
7
1
7
24
24
28
8
1
7
25
28
29
9
1
7
25
29
34
10
1
7
25
29
35
11
1
7
25
40
40
: Floyd
4.3.4
Function Floyd (L[1..n,1..n]):array D[1..n,1..n]

DL
for k 1 to n do
for i 1 to n do
for j 1 to n do
D[i,j] min(D[i, j], D[i, k] + D[k, j])
return D
www.Bshams.ir
www.Bshams.ir
.4
132
(n3 )
Function Floyd (L[1..n,1..n]):array D[1..n,1..n],array P[1..n,1..n]

DL
P
for k 1 to n do
for i 1 to n do
for j 1 to n do
if D[i,j] > D[i,k]+D[k,j]
D[i,j] D[i,k]+D[k,j]
P[i,j] k
return D , P
:
15
4
1 I
30
5
5
15
50
2
15
0 5
50 0 15 5
D0 = L =
30 0 15
15 5 0
0 5 20 10
50 0 15 5
D2 =
30 35 0 15
15 20 5 0
0 5 15 10
20 0 10 5
D4 =
30 35 0 15
15 20 5 0
- 3
0
50
D1 =
30
15
0
45
D3 =
30
15
0
4
P =
0
0
5
0 15 5
35 0 15
20 5 0
5 20 10
0 15 5
35 0 15
20 5 0
0 4 2
0 4 0
1 0 0
1 0 0
? :
?
www.Bshams.ir
Dijkstra
www.Bshams.ir
DYNAMIC PROGRAMMING
133
(chained matrix multiplication)
di1 di
A1 , A2 , ..., An
Ai .
A=A1 A2 .....An
.3.4
5.3.4
.
.

mij
mii = 0
Ai
mii+1 = di1 di di+1
M=(mij )nn
Ai Ai+1 .....Aj
Ai Ai+1
A = (aij )pq B = (bij )qr C = (cij )pr = AB

cij =
q
P
k=1
aik bkj
for i=1 to p do
for j=1 to r do
(pqr)
for k=1 to q do
cij = cij + aik bkj
( Ai Ai+1 ...Ak )(Ak+1 ...Aj )

mij =
0
i=j
minikj1 {mik + mk+1,j + di1 dk dj }
j>i , i=1,2,.....n-1
j=i+s
0
s=0 , i=1,...,n
mi,i+s =
miniki+s1 {mik + mk+1,i+s + di1 dk di+s } i=1,2,.....n-s ,
1 s n-1
www.Bshams.ir
www.Bshams.ir
.4
134
int minmult(int n,const int d[ ],index p[ ][ ]){

index i,j,k,diagonal;
int M[1..n][1..n]
for (i=1;i n;i++)
M[i][[i]=0;
for (diagonal=1;diagonal n-1;diagonal++)
for (i=1;i n-diagonal;i++){
j i+diagonal
m[i][j] = minikj1 {m[i][k] + m[k + 1][j] + d[i 1]d[k]d[j]}
p[i][j]=a value of k that gave the minimum
}
return M[1][n];}
n
X1
s(n s) = (n3 )
s=1
. :
A1 = A135
11
m22
s=0 =
m
33
m
44
m12
s=1 = m23
m34
A2 = B589
A3 = C893
A4 = D334
=0
=0
=0
=0
= d0 d1 d2 = 13 5 89 = 5785
= d1 d2 d3 = 5 89 3 = 1335
= d2 d3 d4 = 89 3 34 = 9078
m13 = {m11 + m23 + 13 5 3, m12 + m33 + 13 89 3}

s=2 = = min{1530, 9256} = 1530
m24 = 1845
www.Bshams.ir
www.Bshams.ir
DYNAMIC PROGRAMMING
135
s=3 =
.3.4
m14 = min{m11 + m24 + d0 d1 d4 , m12 + m34 + d0 d2 d4 , m13 +
m44 + d0 d3 d4 }=2856
6.3.4
.
) . .
.

pi
n
P
i=1
pi =

.

:
node
12
18
20
27
probability 0.2
0.25
0.05
0.1
0.05
34
0.3
35
0.05
.
.
.
34
12
35
20
18
27
10.3+20.25+20.05+30.2+30.1+40.05+40.05 = 2.2
www.Bshams.ir
www.Bshams.ir
.4
n
P
i=1
pi (depth(ci ) + 1)
136
12
6
20
34
18
35
27
10.25+20.2+20.1+30.05+30.3+40.05+40.05 = 2.3
:
.
ci , ci+1 , . . . , cj
cij
. . b
c1n
ck
ci < . . . < ck1
ck+1 < . . . < cj
1)i = j cii = pi
2)j > i
ci < . . . < ck1 < ck < ck+1 < . . . < cj
ck
ci,k1 ck
pk
ck
ck+1,j
ci,k1 + ck+1,j + pk + pi + ... + pk1 + pk+1 + . . .+ pj = cik1 + ck+1j +
j
X
t=i
www.Bshams.ir
pt
www.Bshams.ir
DYNAMIC PROGRAMMING
137
cij = minikj {(cik1 + ck+1j ) +
j
X
.3.4
pt }
t=i
:
ci,i+s =
j = s + i
pi
s = 0, i = 1, ..., n
i+s
P
pt
miniki+s {ci,k1 + ck+1,i+s } +
t=i
1sn1
,1 i ns
n
P1
s=1
(n s)(s + 1) = (n3 ):
: 4
Don
Isabelle
Rulph
Wally
key[1]
key[2]
key[3]
key[4]
p1 = 38
C 1 2
1 38 98
2
3
4
3
8
p2 = 38
p3 = 18
p4 = 18
P 1
1 1
2
1
3
2
4
2
11 7
8 4
5
1
8
1
3
8
8
1
8
Isabelle
Don
Rulph
Wally
www.Bshams.ir
www.Bshams.ir
.4
138

Y = y1 y2 . . . yn , X = x1 x2 . . . xn
7.3.4

, .
.
:
.Yj
= y1 y2 . . . yj ,Xi = x1 x2 . . . xi
LCS=Longest Common Subsequence
LCS[Xi , Yj ] =
1 + LCS[Xi1 , Yj1 ]
M ax{LCS[Xi1 , Yj ], LCS[Xi , Yj1 ]} xi 6= yj , i 6= 0, j 6= 0
LCS-length(X,Y)
1
m length(X)
n length(Y)
for i 1
4
5
6
7
8
9
to
do
for j0
to
do
do
c[0,j]0
for i1
to
for j1
to
do
if (xi = yj ) then
10
c[i,j] c[i-1,j-1]+ 1
11
b[i,j] -
else if (c[i-1,j] c[i,j-1]) then
13
c[i,j] c[i-1,j]
14
b[i,j]
15
else
16
c[i,j] c[i,j-1]
17
b[i,j]
18 return c and b
www.Bshams.ir
c[i,0] 0
12
i = 0 or j = 0
xi = yj , i 6= 0, j 6= 0
//(nm)
www.Bshams.ir
DYNAMIC PROGRAMMING
139
.3.4
Print-LCS(b,X,i,j)
1 if i=0 or j=0 then
2
return
3 if b[i,j]=- then
4
Print-LCS(b,X,i-1,j-1)
Print xi
6 if b[i,j]= then
7
Print-LCS(b,X,i-1,j)
8 else Print-LCS(b,X,i,j-1)
//O(n + m)
.
.
:
X=ABCBDAB
j
i
www.Bshams.ir
Y=BDCABA
LCS=BCBA
yj
xi
www.Bshams.ir
.4
140
8.3.4
.
.( )
2n 1 n
p(i, j)
.
:
p(i, j)
p(i, j) = p p(i 1, j) + q p(i, j 1) , i 1, j 1

p(0, j) = 1
j 1
p(i, 0) = 0
i 1
: Devide and conquer
function p(i,j){
if i=0 then return 1
else if j=0 then return 0
else return p*p(i-1,j)+q*p(i,j-1)
};
p(n, n)
g(i, j) = g(i 1, j) + g(i, j 1) + 1

g(i, 0) = g(0, j) = 0
:
h(i + j, j) = g(i, j)
h(i + j, j) = h(i + j 1, j) + h(i + j 1, j 1) + 1

h(i + 0, 0) = h(0 + j, j) = 0
.
i+j
j
www.Bshams.ir
www.Bshams.ir
DYNAMIC PROGRAMMING
141
4
2n+1
2n 2
n
2n
n
p(n, n)
.3.4
. .
Dynamic programming
function series(n,p){
array p[0..n,0..n]
q=1-p
{ fill from topleft to diagonal }
for s = 1 to n do
p[0,s] = 1 , p[s,0] = 0
for k = 1 to s - 1 do
p[k,s - k] = p * p[k - 1,s - k] + q * p[k,s - k - 1]
{ fill from below diagonal to bottomright }
for s = 1 to n do
for k = 0 to n - s do
p[s + k,n - k] = p * p[s + k - 1,n - k] + q * p[s + k,n - k - 1]
return p[ n , n ]};
(n2 )
3
2
1
0
..1..
....
..1..
0
0
.....1 ........ ........
.
.
.
.
.
.
....
.... .p+pq
....
.....0 ........p
.. ..
.
.
.
.
.
.....
2
..2.. 3
2
....0 ........p p + pq + pq
.
.
.
.
.
.
3 ....0
0
n
www.Bshams.ir
www.Bshams.ir
.4
142
9.3.4
minlenght.
n 1
P[i][A]
v1
vi
V v1
P,
W[i][j]
vi

. f
void travel(int n,const number W,index P,number & minlenght)

{
index i,j,k;
number D[1..n][subset of V-{v1 }];
for (i=2 ;i n;i++)
D[i][]=W[i][1];
for(k=1;kn-2;k++)
for(all subset A V -{v1 }contaning k vertices)
for (i such that i6=1 and vi is not in A){
D[i][A]=minj:vj A (W[i][j]+D[j][A-{vj });
P[i][A]=value of j that gave the minimum ;
}
D[1][V-{v1 }]=min2jn (W[1][j]+D[j][V-{v1 , vj }])
P[1][V-{v1 }]=value of j that gave the minimum;
minleght=D[1][V-{v1 }];
}
www.Bshams.ir
www.Bshams.ir
143
.4.4
4.4
.
:
2

.
:
20
13
20
20
14
19 20
dlg ne
14
12
9
12
13
5
13
2
.
n=2 :
k
lg n
X
n
i=1
2i
2h n
n
(1 ( 12 )lg n )
1
= 2
= n(1 ) = n 1
1
n
1 2
2
,
n-1
.

. , . ( 20 )
n-1+dlg ne-1=n+dlg ne-2
dlg ne -1
www.Bshams.ir
www.Bshams.ir
.4
144
:
.
B &T ( Back Tracking )
5.4

. ( )

B&T
Promissing
: : .
promissing
. (fisible)
.
.
n ,nn
1.5.4
[k,col[k]] k
[i,col[i]] i
k < i
k i
col[i]6= col[k], |col[i] col[k]| 6= i-k

bool promssing (index i) {
index k;
bool switch;
k=1;
switch=true;
while(k < i && switch){
if (col[i]==col[k] k abs(col[i]-col[k])==i-k)
switch=false;
k++;}//{end of while}
return switch;}
.n
www.Bshams.ir
www.Bshams.ir
B &T ( BACK TRACKING )
145
.5.4
void queens (index i){

index j;
if (promissing (i))
if (i==n)
cout<< col[1] through col[n];
else
for(j=1;j <= n;j++){
col[i+1]=j;
queens(i+1);}
}
.
.
wi
n, ..., 2, 1
.
.
f
.
w1 = 3, w2 = 4, w3 = 5
13
w4 = 6
. .
.
w1
w2
w3 1
3
0
8
0
7
13
www.Bshams.ir
12
4
0
0
0
www.Bshams.ir
weightk =
k
P1
i=1
xi wi , totalk =
n
P
.4
146
wi
i=k
void sum-of-subset(index i,int weight,int total){

if (promissing(i)){
if(weight==W)
cout<<x[1] through x[n];
else{
x[i+1] 1
sum-of-subset(i+1,weight+w[i+1],total-w[i+1]);
x[i+1] 0
sum-of-subset(i+1,weight,total-w[i+1]);}
}
}
bool promissing(index i){
return
((weight + total W )&&((weight == W )||(weight + w[i + 1] W ))
}
b
void hamiltonian(index i)
index j;
if(promissing(i))
if(i==n-1)
cout vindex[0] through vindex[n-1];
else
for(j = 2 ;j n ; j++){
vindex[i+1] = j ;
hamiltonian(i+1);}
}
www.Bshams.ir
2.5.4
www.Bshams.ir
B &T ( BACK TRACKING )
147
.5.4
bool promissing(index i){

index j;
bool switch;
if (i==n-1 && !w[vindex[n-1]][vindex[0]])
switch=false;
else if (i > 0 && !w[vindex[i-1]][vindex[i]])
switch=false;
else {
switch=true;
j=1;
while(j <i && switch){
if(vindex[i]==vindex[j])
switch=false;
j=j+1;}
}
return switch;}
m-coloring b

m b n
3.5.4
. b
.
void m-coloring (index i){

int color;
if (promissing(i))
if (i == n)
cout << vcolor[1] through vcolor[n]
else
for (color = 1; color <= m; color++) {
volor[i+1] = color;
m-coloring(i+1);}
}
www.Bshams.ir
www.Bshams.ir
.4
148
bool promissing (index i) {

index j;
bool switch = true;
j = 1;
while (j < i && switch) {
if (W[i][j] && vcolor[i] == vcolor[j])
switch = false;
j++; }
return switch;
}
Branch and Bound (B&B)

Back Tracking
6.4
( )
Bound
B&B
f .

,

b
f . . ,
.
www.Bshams.ir
BFS
www.Bshams.ir
Exploring graphs .5
(Depth First Search) DFS
procedure DF-Search(G=< N,A > )
for each v N do mark[v] not-visited
for each v N do
if mark[v] 6= visited then
DFS(v)
procedure DFS(v)
{Node v has not previously been visited}
mark[v] visited
for each node w adjacent to v do
if mark[w] 6= visited then
DFS(w)
www.Bshams.ir
1.5
www.Bshams.ir
.5
150
1
2
1.
2.
DFS(1)
= mark[1]=visited , mark[2] 6= visited

DFS(2)
3.
DFS(3)
4.
DFS(6)
5.
6.
DFS(5) .
visit
DFS(4)
7.
DFS(7)
8.
DFS(8)
:
www.Bshams.ir
DF-Search
A
(|N | + |A|)
www.Bshams.ir
(DEPTH FIRST SEARCH) DFS
151
.1.5
:

.

. b
.
DF-Search
1
6
2
3
6I
6
w 6
4
6
7
R ?
8
I
-

2
?
3
K
1
~
?
q 8
?
7
1. DFS(1)
2.
DFS(2)
3.
4.
DFS(3)
DFS(4)
5.
DFS(8)
6.
DFS(7)
7. DFS(5)
8.
www.Bshams.ir
DFS(6)
5
?
6
www.Bshams.ir
.5
152
1.1.5
Procedure DFS2(v)
p empty-stack
Mark[v]visited
push v on to p
while p is not empty do
while there exites a node w adjacent to top(p)
such that mark[w]6=visited do
mark[w] visited
push w on to p{w is the new top(p)}
pop(p)
(Breath First Search) BFS

procedure
2.5
BF-Search(G)
for each v N do Mark[v] not-visited

for each v N do
if Mark[v] 6= visited then
BFS(v)
procedure BFS(v)
Q empty-queue
mark[v] visited
enqueue
While
into
Q is not empty
do
u first(Q)
dequeue
for
from
each node w adjacent to u
do
if Mark[w] 6= visited then

Mark[w] visited
enqueue

visit
w into
O(|N | + |A|)
www.Bshams.ir
www.Bshams.ir
TOPOLOGICAL SORT .3.5
153
node visited
1.
2.
3.
4.
5.
6.
7.
8.
1
2
3
4
5
6
7
8
16 , 2
, 3, 4
| {z }
26 , 3, 4, 5
,6
|{z}
36 , 4, 5, 6
46 , 5, 6, 7
,8
|{z}
56 , 6, 7, 8
66 , 7, 8
76 , 8
86
Topological Sort
3.5
b )
(
:
) . b
b b (
. f .
.
. , f
.
(1)
(|N | + |A|)
DFS

: .
*3
-6
s 2?
-4
1,3,2,4,6,5
www.Bshams.ir
j
5
3
www.Bshams.ir
.5
154
Bellman Ford
Bellman
4.5
Ford

.
.
.
G=< V, E >
V[G]
IN IT ILIZE SIN GLE SOU RCE(G, s)

1.
for each vertex
v V [G]
2.
d[v]
3.
[v] N IL
4.
do
d[s] 0
RELAX(u, v, w)
1.
if
d[v] >d[u]+w[u,v]
2.
d[v] d[u] + w[u, v]
3.
[v] u
then
BELLM AN F ORD(G,w,s)
1.
INITIALIZE-SINGLE-SOURCE(G,s)
2.
for
3.
i 1
for
4.
5.
to (|V [G]|-1)
each
edge
(u, v) E(G)
do
RELAX(u,v,w)
for
each
edge
(u, v) E(G)
6.
if d[v] > d[u]+w(u,v)
7.
return FALSE
8.
do
return TRUE
then
//(|V ||E|)
www.Bshams.ir
www.Bshams.ir
DAG
155
.5.5
5
r
^x
6 3 I-2 6
-3
8
7
s 0y2
-4
?
R
j
7
-
9
y
z
R x
6
-2
6> Y
6
8
-3
s 0 y2
7
-4
R
7s ?
-
7
9
y
z
r
(b)
(a)
5
r
^ x
6 -2 4
6 I 7 6
8
-3
7
2
i
s 0
? -4 w
-2
7s 7
9
z
y
5
r
^ x
4
2
6 3 I -2 6
8
-3
7
s 0 y2
-4
7 s ?
R
2
7
9
y
z
(d)
(c)
5
x
4
63 I-2 6
8
-3 7
s 0 y2 ? -4R
q
- -2
7 7
y 9
z
(e)
r
2
DAG
5.5
DAG-Shortest-Path(G,w,s)
1.
Topologically sort the vertics of G
2.
initialize-single-source(G,s)
3.
for each vertex u,taken in Topologically sorted order do

for each vertex vAdj[u] do
RELAX(u,v,w)
www.Bshams.ir
www.Bshams.ir
.5
156
(|V | + |E|)
6
1
x
y
t
sz
~
2- 7- -1- -2
-
*
>
>
4
: .
s
5- 0
3
2
6
r
50
t
2-
3
1
x
y
z
7-s -1- -2-~
*
4

3
r
s
5- 0
6
t
22
>
1
y
sx
~z
7 - 6 -1- -2-

4

3
r
s
50
6
t
22

1
z
y
s
-1-2
6
4
*
4
>
2
x
7 -s
6
1
y
z
s
t
~
s x -1
-2
2575
6
4
0
2
3
3
4 3
2
3
6
1
x
r
y
s
jz
t
j
5 - 0 2- 2 7 - 6 -1- 5 -2- 3
3
>

4
3
2
6
1
x
r
z
s
y
t
s
-2 s
-127 5- 0
6
2
5 -3
*
*
3
4
3
2
6
www.Bshams.ir
www.Bshams.ir
LOOP INVARIANT
1.6
? LOOP INVARIANT
.
,
Loop Invariant .
Loop Invariant
Loop Invariant
,
.
LOOP INVARINT
www.Bshams.ir
Loop Invariant
www.Bshams.ir
.6
158
...
// the Loop Invariant must be true here
while ( TEST CONDITION ) {
//top of the loop
...
//bottom of the loop
//the Loop Invariant must be true here
}
// Termination + Loop Invariant Goal
...
Loop Invariant
Loop Invariant
Loop Invarint
Loop Invariant
: .
. :Pre-condition
:Post-condition

Loop Invariant
:Loop
Variant
. Pre-condition 1
. Pre-condition Loop Invariant 2
. Loop Invariant Loop Variant 3
. Loop Invariant 4
www.Bshams.ir
www.Bshams.ir
LOOP INVARIANT .1.6
159
Loop Invariant
. Post-condition
. Loop Variant 6
loop Invariant

2 . 14
4
. 5
.
:
n 1
(1
1. int sum=0;
2. int k=0;
3. while(k < n){
4.
k++;
5.
sum+=k;
6.
........................................................................................................
precondition : sum = 0, k = 0
postcondition : sum =
n
P
i=1
loop invariant : sumk =
INITIALIZATION:
www.Bshams.ir
k
P
i=1
www.Bshams.ir
.6
precondition
160
loop invariant
k = 0 sum =
0
P
i=1
i = 0 = sum
MAINTENANCE:

j loop invariant
kj+1
sumj =
j=1
j+1
j , kj+1 = kj + 1
sumj+1 = sumj + kj+1 = (
kj
P
j=1
j) + kj+1 = (
kj
P
j=1
j) + kj + 1 =
kj+1
j=1
TERMINATION:
loop invariant
k=n
postcondition
:
sum =
n
P
i=1
i postcondition
: (2
1. int factorial(n){
2.
i =1;
3.
fact =1;
4.
while(i ! = n){
5.
i++;
6.
fact=facti; }
7.
return fact;
8.
.............................................................................................................
precondition : n 1
www.Bshams.ir
www.Bshams.ir
LOOP INVARIANT .1.6
161
loop invariant : f act = i !

postcondition : f act = n !
........................................................................................
INITIALIZATION:
i = 1 f act = 1! = 1 f act = i !
MAINTENANCE:
f act0 = j 0 ! , j = j 0 + 1 , f act = f act0 j
f act = j 0 ! j = j 0 (j 0 + 1) = (j 0 + 1)! = j! f act = j!
TERMINATION:
i = n , f act = i! f act = n! postcondition
precondition
, :
n 1
(3
:
1. int gcd(int m ,int n){

2.
int mprime = m;
3.
int nprime = n;
4.
while(mprime ! =nprime){
5.
if(mprime > nprime)
6.
mprime - = nprime;
7.
8.
9.
10.
else
nprime - = mprime;}
return mprime;
}
.....................................................................................................
precondition : m, n Z +
www.Bshams.ir
www.Bshams.ir
.6
162
loop invariant : gcd[m, n] = gcd[mprime, nprime]
postcondition : gcd[m, n] = mprime

...........................................................................
INITIALIZATION:
mprime = m
nprime = n gcd[m, n] = gcd[mprime, nprime]
MAINTENANCE:
gcd[m, n] = gcd[mprimei , nprimei ]
if (mprimei > nprimei ) :

mprimei+1 = mprimei nprimei ,
nprimei+1 = nprimei
gcd[mprimei+1 , nprimei+1 ] = gcd[mprimei nprimei , nprimei ] =

gcd[mprimei , nprimei ] = gcd[m, n]
else :
nprimei+1 = nprimei mprimei ,
mprimei+1 = mprimei
gcd[mprimei+1 , nprimei+1 ] = gcd[mprimei , nprimei mprimei ] =

gcd[mprimei , nprimei ] = gcd[m, n]
TERMINATION:
mprime = nprime gcd[m, n] = gcd[mprime, nprime] =
gcd[mprime, mprime] = mprime postcondition
en
1. double TaylorExp(double n ,int k){
www.Bshams.ir
(4
www.Bshams.ir
LOOP INVARIANT .1.6
163
2.
double result =1;
3.
int count =0;
4.
int denom=1;
5.
while (count<k){
6.
count ++;
7.
denom=count;
8.
result+=pow(n,count)/denom;
9.
10.
}
return result;
11.
......................................................................................................
precondition : n Z
k N, k > 0
2
loop invariant : result = 1 + n + n2! + +

2
ncount
count!
, denom = count!
postcondition : 1 + n + n2! + + nK! = result

...............................................................
K
INITIALIZATION:
count = 0,
denom = 1,
result = 1 n0! = 1 = result
MAINTENANCE:
2
resulti = 1 + n + n2! + +
ncounti
counti !
, denomi = counti !
denomi+1 = denomi counti+1 = (counti !)counti+1 = (counti+1 )!,

2
n
n
resulti+1 = resulti + (count
= 1 + n + n2! + + ncounti ! + (count
i+1 )!
i+1 )!
counti+1
counti
TERMINATION:
2
count = k result = 1 + n + n2! + +
www.Bshams.ir
nK
K!
postcondition
counti+1
www.Bshams.ir
.6
164
loop invariant
(5
:
function add(y,z)
comment
return y + z, where y,z N
1.
x := 0 ; c := 0 ; d := 1 ;
2.
while(y > 0 ) (z > 0 ) (c > 0 )do
3.
a := y mod 2;
b :=z mod 2;
4.
if a b c
then x : = x+d ;
5.
c := (a b) (b c) (a c);
6.
d :=2 d;
y := b y /2 c;
z:=b z /2 c;
7.
return(x)
.......................................................................................................
loop invariant : (yj + zj + cj )dj + xj = y0 + z0
z0 y0
z y
x0 = 0, c0 = 0, d0 = 1
(yj + zj + cj )dj + xj = (y0 + z0 + 0) 1 + 0 = y0 + z0

:
loop invariant
(yj + zj + cj )dj + xj = y0 + z0
:
aj+1 = yj mod 2
, bj+1 = zj mod 2
yj+1 = byj / 2 c , zj+1 = bzj /2c , dj+1 =2dj
bj+1 aj+1
cj+1
www.Bshams.ir
cj+1
bj+1 aj+1
, cj
cj
www.Bshams.ir
LOOP INVARIANT .1.6
165
cj+1 = b(aj+1 + bj+1 + cj ) / 2 c
xj+1
xj+1
dj
xj
aj+1 bj+1 cj
xj+1 = xj + dj ((aj+1 + bj+1 + cj ) mod 2)
(*)
2b n /2c + ( n mod 2)= n
n N
(*)
loop invariant
(yj+1 + zj+1 + cj+1 )dj+1 + xj+1 =

(byj /2c + bzj /2c + b(yj mod 2 + zj mod 2 + cj )/2c) 2dj + xj +
dj ((yj mod 2 + zj mod 2 + cj ) mod 2) = (byj /2c + bzj /2c) 2dj
+xj + (b(yj mod 2 + zj mod 2 + cj )/2c) 2dj +
dj ((yj mod 2 + zj mod 2 + cj ) mod 2) () (byj /2c + bzj /2c) 2dj
+xj + dj (yj mod 2 + zj mod 2 + cj ) = byj /2c 2dj +
dj (yj mod 2) + bzj /2c 2dj + dj (zj mod 2) + cj dj + xj ()
(yj + zj + cj ) dj + xj = y0 + z0

,
:
loop invarint
(yk + zk + ck )dk + xk = y0 + z0

bz/2c by/2c
yk = zk = ck = 0 xk = y0 + z0
www.Bshams.ir
z y
www.Bshams.ir
.6
166
z y
loop invariant (6
:
function multiply(y,z)
comment
return y z , where y , z N
1.
x :=0
2.
while (z >0)do
3.
x :=x + y (z mod 2);
4.
y := 2 y;
5.
return(x)
z :=b z/2 c;
........................................................................................................
loop invariant : xj + yj zj = y0 z0
x0 =
0 , z0 y0 z y
:
x0 = 0 xj + yj zj = x0 + y0 z0 = y0 z0
loop invariant
xj + yj zj = y0 z0
:
xj+1 = xj + yj (zj mod 2)
yj+1 = 2 yj
zj+1 = bzj /2c
xj+1 + yj+1 zj+1 = xj + yj (zj mod 2)+2yj (bzj /2c) =

xj + yj ((zj mod 2)+2bzj /2
www.Bshams.ir
c))()
=
xj + yj zj = y0 z0
www.Bshams.ir
LOOP INVARIANT .1.6
167
z
:
zk = 0, xk + yk zk = y0 z0 xk + yk zk = xk + yk 0 = xk = y0 z0

loop invariant
(7
function divide(y,z)
comment
return q, r Nsuch that y = qz+r
and r < z , where y , z N

1.
r :=y; q :=0; w :=z;
2.
while w y do w :=2 w;
3.
while w > z do
4.
q :=2 q; w := b w /2 c ;
5.
if w r then
6.
r :=r - w; q :=q+1;
return (q , r)
7.
.......................................................................................................
loop invariant : qj wj + rj = y0 , rj < wj
r = y0
q=0
y0
r = y0
, q = 0 qj wj + rj = 0 + y0 = y0
: ,
www.Bshams.ir
loop invariant
www.Bshams.ir
.6
168
qj wj + rj = y0 , rj < wj
w>z
wy
, ,
bw/2c ,
w = z
w/2
bw/2c
qj+1 = 2qj
, wj+1 = bwj /2c
j + 1
w>z

wj+1 > rj
1:
rj+1 = rj
, qj+1 wj+1 + rj+1 =
2qj bwj /2c + rj = 2qj (wj /2) + rj = qj wj + rj = y0

2:
wj+1 > rj rj < wj+1
, rj+1 = rj
rj+1 < wj+1
www.Bshams.ir
(a)
(b )
www.Bshams.ir
LOOP INVARIANT .1.6
169
wj+1 < rj
1:
rj+1 = rj wj+1 = rj bwj /2c, qj+1 =

qj+1 + 1 = 2qj + 1 qj+1 wj+1 + rj+1 =
(2qj + 1)bwj /2c + rj bwj /2c =
qj wj + bwj /2c + rj bwj /2c = y0
2:
wj+1 < rj , rj+1 = rj wj+1
(1)
rj+1 < wj+1 (1)

rj+1 wj+1
,
:
rj wj+1 wj+1 rj 2wj+1 rj 2bwj /2c rj wj

loop invariant
rj+1 < wj+1 : j

j+1 loop invariant
,
.

.wk
qk z0 + rk = y0
, rk < z0
= z0
loop invariant
loop invariant
(8
function power(y,z)
comment
www.Bshams.ir
return yz , where y R,z N
1.
x :=1;
2.
while z > 0 do
3.
if z is odd then x := x.y;
4.
z :=bz /2 c;
5.
y :=y2 ;
6.
return (x)
www.Bshams.ir
.6
170
loop invariant : xj yj zj = y0 z0
loop
x0 =
1 , z0 y0 z y
:
invariant
xj yj zj = y0 z0
j+1 j
loop invariant
xj yj zj = y0 z0 ,
2
z xj+1 = xj yj , zj+1 = bzj /2c, yj+1 = yj
z xj+1 = xj ,
zj+1 = bzj /2c,
yj+1 = yj 2
:
2bzj /2c
z :
xj+1 yj+1 zj+1 = xj yj (yj )
=
2(zj 1)/2
= xj yj zj = y0 z0
xj yj (yj )
2bzj /2c
z :
xj+1 yj+1 zj+1 = xj (yj )
=
2(zj )/2
= xj yj zj = y0 z0
xj (yj )
z=0
xk yk zk = y0 z0 , zk = 0 xk = y0 z0
. ,
www.Bshams.ir
www.Bshams.ir
LOOP INVARIANT .1.6
171
loop invariant
(9
A[1 n]
function sum(A)
comment
return
1.
s :=0;
2.
for i := 1 ton do
3.
s :=s+A[i]
4.
return (s)
Pn
i=1
A[i]
.....................................................................................................
loop invariant : sj =
j
X
A[i]
i=1
loop
j = 0 s = s0 =
,
0
X
i=1
invariant
A[i] = 0
loop invarint
:
sj+1 = sj + A[j + 1] = (
j
X
i=1
A[i]) + A[j + 1] =
j+1
X
j+1
A[i]
i=1
j = n s = sn =
n
X
A[i]
i=1
.
www.Bshams.ir
www.Bshams.ir
.6
172
(10
Horner an xn + an1 xn1 + a1 x + a0

:
A[i] = ai
loop invariant
f or
A[0..n]
0in
all
function Horner(A , n)
Pn
comment return i=0 A[i].xi
1.
v := 0
2.
for i := n downto 0 do
3.
v := A[i]+v.x
return (v)
4.
..................................................................................................
loop invariant : vj =
n
X
A[i]x ij
i=j
for
j =n+1 v =
n
X
i=n+1
: ,
vj = A[j] + vj+1 .x vj+1 =

n
P
i=j+1
A[i]xij1 =
n
P
i=j+1
j=0v=
n
P
i=0
vj A[j]
x
loop invariant
loop invariant
j=n+1
A[i]xi(n+1) = 0
loop
n
P
invariant
n
P
A[i]xij A[j]
i=j
A[i]xij
i=j+1
A[i]xi(j+1)
A[i]xi
j=0
www.Bshams.ir
www.Bshams.ir
(AMORTIZED ANALYSIS)
173
(Amortized Analysis)
.2.6
2.6
. ,

.

.

.
.
(Aggregate
Analysis)
(Acconting
Method)
(Potential
Method)
(Aggregate Analysis)

1.2.6
, .
.
T (n) f,
T (n)
n
,
.
.
.
.
www.Bshams.ir
www.Bshams.ir
.6
174
push(S, x)
O(1)
: 1
. ,

push
O(1)
pop(S)
. 1
n , pop
multipop(S, k)
S
TRUE Stack-Empty
FALSE
Multipop(S, k)
1. while
not
Stack-Empty(S)
2.
do pop(S)
3.
k k1
multipop
pop
and
push
k 6= 0
,
.
.
O(n2 )
O(n)
multipop
O(n)

.
, .
push
O(n)
multipop
multipop
pop

, (multipop )
.
push
pop
, .
stack
www.Bshams.ir
www.Bshams.ir
175
O(n)
multipop
, .
O(n)
n
pop , push n
= O(1)
.2.6
.
.
O(1)
b
.
.
2 k-bit
: 2
A[0]
length[A]=k
A[0..k 1]
: .
A[k 1]
INCREMENT(A)
1: i 0
2: while
i < length[A] and A[i] = 1
3:
do A[i] 0
ii+1
4:
5:
if
i < length[A]
then
6:
A[i] 1
1 )

INCREMENT
INCREMENT n (k) (
.
O(nk)
A[1]
.
.
b4c
A[2]
b 2ni c , A[i]
A[0]
O(n)
n
b2c
INCREMENT
i = 0, 1, 2, , blog2 c
n
A[i]
i > blogn2 c
Incrementing a Binary Counter
www.Bshams.ir
www.Bshams.ir
.6
176
blg nc
X
i=0
c<n
i
O(2n)
n
1
2i
i=0
= O(1)
= 2n
(Accounting Method)
2.2.6
,

.
, .

f .
.
.
.

,
.
i
C
Ci

: ,
n
X
Ci
i=1
n
X
Ci
i=1
n
X
i=1
Ci
n
X
Ci
i=1
.

.
www.Bshams.ir
www.Bshams.ir
177
.2.6
:1
Push
: 1
Pop
: 2
Multipop : 3
Push
: 2
Pop
: 0
Multipop : 0
multipop
( ) 1

b
.
.
b
b

b
multipop
pop
pop
. .

O(n)
multipop
pop
push
.
f .

:2
.
1 . 2

1 , .
,

O(n)
.
.
www.Bshams.ir
www.Bshams.ir
.6
178
(Potential Method)
3.2.6
,
,
b
.
.

D0
1
:
i = 1, 2, , n
Ci = i
Di =
=(Potential Function) Di
: Di RealN umber
C0 = Ci + (Di ) (D
i
i1 )
n
n
n
n
P
P
P
P
[Ci + (Di ) (Di1 )] =
Ci +
(Di ) (Di1 )
Ci =
i=1
i=1
n
P
i=1
i=1
i=1
Ci + (Dn ) (D0 )
(Dn ) > (D0 )
. .
(Di ) (D0 )
(D0 ) = 0
(Di ) 0
.
:
(Di ) (0) > 0
for all i
.
.
.
www.Bshams.ir
www.Bshams.ir
179
: .
(Di ) =
D0 =
.2.6
: 1
(D0 ) = 0

s , i 1
(Di ) 0 = (D0 )

(Di1 ) = s
i push
i pop
(Di ) (Di1 ) = (s + 1) s = 1
Ci = Ci + (Di ) (Di1 ) = 1 + 1 = 2 O(1)
:
(Di ) (Di1 ) = (s 1) s = 1
Ci = Ci + (Di ) (Di1 ) = 1 1 = 0 O(1)
:
k = min(s, k) (Di ) (Di1 ) = (s k ) s = k

Ci = Ci + (Di ) (Di1 ) = k 0 k 0 = 0 O(1)
.
O(n)
multipop
(Di ) = bi
:2
(Di ) 0
ti =
Ci = ti + 1
www.Bshams.ir
www.Bshams.ir
.6
if
bi = 0 bi1 = k = ti
if
bi > 0 bi = bi1 ti + 1
bi bi1 ti + 1
(Di ) (Di1 ) bi1 ti + 1 bi1 = ti + 1

Ci0 = Ci + (Di ) (Di1 ) ti + 1 ti + 1 = 2 Ci0 2
n
P
i=1
www.Bshams.ir
Ci0 2n O(n)
180

Algorithm HajSeyedJavadi

Caricato da

Informazioni sul documento

Copyright

Formati disponibili

Condividi questo documento

Condividi o incorpora il documento

Opzioni di condivisione

Hai trovato utile questo documento?

Questo contenuto è inappropriato?

Copyright:

Formati disponibili

Algorithm HajSeyedJavadi

Caricato da

Copyright:

Formati disponibili

www.Bshams.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (Master Theorem) 3.2

. . . . . . . . . . . . . . . . . . . . Pigeon hole Sort

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (Sparse array) 1.3

Binomial Heap 4.3

Max Binomial Tree 2.4.3

. . . . . . . . . . . . . . . . . . . . . . . The Merge Of Max Binomial Trees 3.4.3

Max Binomial Heap

Max Fibonacci Tree 2.5.3

&T ( Back Tracking )

147 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . m-coloring b 3.5.4

Branch and Bound (B&B) 6.4

First Search) DFS

. . . . . . . . . . . . . . . . . . . . . . . . . . (Breath First Search) BFS

O(f (n)) = {g:N R+ |c R+ , n N g(n) cf (n)}

O(n2 ) = {n2 , n lg n, f (n), ....}

(f (n)) = {g : N R+ |d > 0 n N f (n) dg(n)}

f (n) < dg(n) = f (n) dg(n)

f2 (n) (f1 (n)) f1 (n) O(f2 (n))

(f (n)) = O(f (n)) (f (n))

O(n2 ) = {n2 , n2 + 4, nlogn, ...}

(n2 ) = {n2 , n2 + n, 2n2 , ...}

o(f (n)) = {g : N R+ | c > 0 n N g(n) cf (n)}

n o(2n) c > 0 n N n c2n

(f (n)) = {g : N R+ | c > 0 n N g(n) cf (n)}

d0 > 0 g(n) (f (n))

n M1 g(n) d0 f (n) (g(n) (f (n)))

2 f (n) d0 f (n) = d0 2d0 = d0 0

1)f (n) + g(n) O(M AX{f (n), g(n)})

f (n) M AX{f (n), g(n)}

g(n) M AX{f (n), g(n)} =

f (n)+g(n) 2M AX{f (n), g(n)} = f (n)+g(n) O(M AX{f (n), g(n)})

n! = ( ne )n 2n = log n! = log( ne )n + log 2n

= log n! = n log ne + log 2n

= log n! = n log n log en + log 2n

= log n! = n log n + log e2nn

logn! = logn + log(n 1) + . . . + log 1 logn + logn + . . . + logn nlogn

nlogn 2logn! 2nlog

logn logn! nlogn

o(f (n)) O(f (n))\(f (n))

g(n) O(f (n))

g(n) O(f (n))

g(n) o(f (n))

g(n) O(f (n))

g(n) o(f (n))

o(f (n)) O(f (n))\(f (n)) = O(f (n))\(O(f (n))) (f (n))

M = 12 , n 13 g(n) n = g(n) O(n)

g(n) (n) c >

g(n) o(n) c > 0

g(n) cn, c = 12 = g(n) n2 F

f (n) O(f (n)) c > 0

g(n) O(f (n))

g(n) O(f (n))

= g(n) c1 f (n) c1 c2 h(n) = g(n) ch(n)

f (n) O(f (n))

, f (n) (f (n)) f (n) (f (n))

g(n) (f (n)) = g(n) O(f (n)), g(n) (f (n)) = f (n) (g(n)),

, g(n) O(h(n)) f (n) O(h(n))

, g(n) (h(n)) f (n) (h(n))

f (n) O(g(n)) = 2f (n) O(2g(n) )

f (n) O(g(n)) = (f (n))k O(g(n)k )

f (n) cg(n) = (f (n))k ck (g(n)k ) = (f (n))k O(g(n)k )

f (n) O(f 2 (n))

f (n) = n1 = f 2 (n) = n12 = f (n) 6 O(f 2 (n))