Models of Computation, HW3 Solutions

Standard mathematical notation: implies that,  there exists, for each, such that.

Problem 1: Prove the following are not regular

A (122/4b): L = {anblak : k  n+l }

Consider L = {anblak : k  n+l }

By construction, L  L = {a*b*a*} and L  L = 
 L = {w : w  L, w  {a*b*a*}} = Lc  {a*b*a*} (Lc == L complement)

The language {a*b*a*} is regular (since a*b*a* is a regular expression).

Furthermore, since regular languages are closed under complement and intersection, we have:
If L is regular then Lc is regular then Lc  {a*b*a*} = L is regular.

Therefore, if L is not regular then L is not regular either. It suffices to show that Lis not regular.
____________

Assume (for contradiction) that L is regular.

Since Lis regular and infinite we have from the pumping lemma that there exists m such that:
w  L with |w|  m,  x, y, z such that:
|xy|  m, |y| = k > 0, w = xyz and xyiz  L  i  0.

Let w = xyz = amba(m+1)  L. By construction, y = ak for some k  [1, m].

Consider xy2z  L
a(m+k)ba(m+1)  L
 m+k+1 = m+1 (by definition of L)
k=0

but k>0 (from pumping lemma)

which is a contradiction
 L is not regular.
 L is not regular.

B (122/4g): L = {wwwwR : w  {a,b}*}

Note:  w, |w| = |wR|

 u (= wwwwR) L,  c  0 such that |u| = 4|w| = 4c

Assume (for contradiction) that L is regular.

Since L is regular and infinite we have from the pumping lemma that there exists m such that:
 u  L with |u|  m,  x, y, z such that
|xy|  m, |y| = k > 0, u = xyz and xyiz  L  i  0.

Let w = amb (u = xyz = ambambambbam). By construction, y = ak for some k  (0,m).

Since |xy|  m and xy is a substring am
 y = ak
Consider u = xy2z  L
u = a(m+k)bambambbam  L
there exists string wsuch that u = www(w)R
We can assume k = 4n, since if k  4n for some n, then |u|  4c and u L
 We can further assume n > 0, since if n = 0, k = 0, and  |y| = 0, which is a contradiction
Since |u| = 4m+4, |u| = 4m+4+k, |w| = m+1+n
n 1, 3n > 1
m + 1 + n < m + 3n + n = m + 4n = m + k
 w = a(m + n + 1) (first m+n+1 characters of u)  (first m+k characters of u) = a(m+k)
 u = wwwwR = a4(m + n + 1)
which is a contradiction, since u is known to contain 4 ‘b’s
 L is not regular.

Problem 2: Are the following regular?

A (122/5b): L = {an : n is not prime}

No.
Consider Lc = {an : n is prime}. (Lc is the coplement of L).

From the closure of regular languages under complement we have:

If L is regular then Lc is regular.

Therefore, if Lc is not regular then L is not regular either. It suffices to show that Lc is not regular.

Assume (for contradiction) that Lc is regular.

Since Lc is regular and infinite we have from the pumping lemma that the exists m such that:
 w  Lc with |w|  m,  x, y, z such that:
|xy|  m, |y| = k > 0, w = xyz and xyiz  Lc  i  0.

Let p > m be prime ; then w = xyz = ap  Lc

By observation, y must be ak
Consider xy(p+1)z  Lc
a(p + kp)  Lc
 (p+kp) is prime (by definition of Lc)

but (p+kp) = p*(k+1)

 (p+kp) is not prime
which is a contradiction
Lc is not regular

B (122/5d): L = {an : n = 2k for some k }

No.
Assume (for contradiction) that L is regular.

Since L is regular and infinite we have from the pumping lemma that there exists m such that:
 w  L with |w|  m,  x, y, z such that:
|xy|  m, |y| = k > 0, w = xyz and xyiz  L  i  0.

Let c = 2j > m ; then w = xyz = ac  L

By observation, y must be ak
Consider xy2z  L
a(c + k)  L
 q  2q = c+k (by definition of L)

but k  m < c,
 2j = c < c+k < 2c = 2*2j = 2(j+1)
  q  2q = c+k
which is a contradiction
 L is not regular

Problem 3: Which of the following are regular?

A (123/9c): L = {anbl : n is an integral multiple of l }

No.
We can use the pumping lemma.

Assume (for contradiction) that L is regular. By definition,

 m   w  L, |w|  m,  x, y, z  |xy|  m, |y| = k > 0, w = xyz and xyiz  L  i  0.
Let w = xyz = ambm  L (m = 1*m)
Since |xy|  m, xy  am
 y = ak
Consider xy2z  L
a(m+k)bm  L
 n  (m+k) = n*m (by definition of L)

but k<m
 1m = m < m+k < m+m = 2m
 n  (m+k) = n*m
which is a contradiction
 L is not regular

B: L = {anbl : (n-l) is a multiple of 3 }

Yes.
The following equivalent statements are used to simplify the problem:
(n-l) is a multiple of 3
 x  n-l = 3x + 0
n-l  0 (mod 3)
n  l (mod 3)
The last of these has 3 cases:
n  l  0 (mod 3) n  l  1 (mod 3) n  l  2 (mod 3)
Note that the extended regular expression { (cx)*cy : x > y } represents { cn : n  y (mod x) }
(this is only an extended regular expression when x and y are left as exponents of c; when
expanded out to { (cccc…c)*ccc…c } , it is a normal regular expression)
Since anbl is simply a concatenation of two such expressions, the union of the 3 cases is:
L = L( ((aaa)* (bbb)*) + ((aaa)*a (bbb)*b) + ((aaa)*aa (bbb)*bb) )
 L is regular
note: the above can also be simplified to: (aaa)*( + ab + aabb)(bbb)*

C (123/9g): L = {anbl : |n-l| = 2 }

No.

We use the pumping lemma.

Assume (for contradiction) that L is regular. By definition,

 m   w  L, |w|  m,  x, y, z  |xy|  m, |y| = k > 0, w = xyz and xyiz  L  i  0.
Let w =xyz = amb(m-2)  L ( |(m) – (m – 2)| == 2)
Since |xy| < m, xy  am
 y = ak
 Consider xy2z  L
a(m+k)b(m-2)  L
| (m+k) – (m–2) | = | k + 2 | = k + 2 = 2 (by definition of L)
k=0
but k>0 (by definition of regular, above)
which is a contradiction
 L is not regular

D: L = {anbl : n + l = 100 }
Yes.
L is finite (101 elements), and all finite languages are regular; therefore L is regular.

Problem 4: Find a Context-Free Grammar for the following:

A (134/8c): L = {anbmck : k = n + m }
Let G be the grammar with productions:
S  aSc | B
B  bBc | 

Claim: L(G) = L
Proof:
Consider the following derivation:
S * anScn  anBcn * anbmBcmcn  anbmc(n + m)
(where the first * applies S  aSc n times, the second B  bBc m times)
Since all words in L(G) must follow this pattern in their derivations, it is clear that L(G)  L

Consider w  L, w = anbmc(n+m) for some n, m  0

The derivation S * anScn  anBcn * anbmBcmcn  anbmc(n + m) clearly produces w for any n, m.
 L  L(G)
 L  L(G)
 G is a CFG for L
B (134/8c): L = {anbmck : k  n + m }
Let G be the grammar with productions:
S  EcC | aAE | AU
A  aA | 
B  bB | 
C  cC | 
E  aEc | F
F  bFc | 
U  aUc | V
V  bVc | bB

Note: L(E) = {anbmck : k  n + m } (from 4(a))

Let L1 = {anbmck : k  n + m }, L2 = {anbmck : k > n + m }

Claim: L(G) = L = L1  L2

Proof:
Consider the leftmost productions following S  EcC
S  EcC * anbmc(n+m)cC * anbmc(n+m)cckC  anbmc(n+m)cck = anbmc(n+m+k+1)
(where the first * follows from 4(a), the second from k repetitions of { C  cC })
Since (n+m) < (n+m+k+1), L(EcC)  L2  L

Consider the leftmost productions following S  aAE

S  aAE * aakAE  aakE * aakanbmc(n+m)c = a(n+k+1)bmc(n+m)
 (where the first * follows from k repetitions of { A aA }, the second from 4(a))
Since (n+m+k+1) > (n+m), L(aAE)  L1  L

Consider the leftmost productions following S  AU

S  AU * akAU  akU * akanUcn  akanVcn * akanbmVcmcn  akanbmbBcmcn
* akanbmbbjBcmcn  akanbmbbjcmcn = a(k+n)b(m+j+1)c(m+n)
(where the first, second, third and fourth * are k, n, m, and j repetitions of
{ A  aA }, { U  aUc }, { V  bVc }, and { B  bB }, respectively)
Since ((k+n)+(m+j+1)) > (m+n), L(AU)  L1  L

 L(G) = L(S) = L(EcC)  L(aAE)  L(AU)  L1  L2 = L

Consider w L1, w = anbmc(n+m+k) for some n, m, k, with k > 0

Then w has the leftmost derivation:
S  EcC * anbmc(n+m)cC * anbmc(n+m)cc(k-1)C  anbmc(n+m)cck(k-1) = anbmc(n+m+k) = w
 L2  L(G)

Consider w  L2, w = a(n+j)b(m+k)c(n+m) for some j, k, m, n, with j+k > 0

If k = 0, w has the leftmost derivation:
S  aAE * aa(j-1)AE  aa(j-1)E * aa(j-1)anEcn  aa(j-1)anFcn * aa(j-1)anbmFcmcn
 aa(j-1)anbmcmcn = a(j+n)b(0+m)c(m+n) = w
 L1 (k = 0)  L(G)
If k > 0, then w has the leftmost derivation:
S  AU * ajAU  ajU * ajanUcn  ajanVcn * ajanbmVcmcn  ajanbmbBcmcn
 ajanbmbb(k-1)Bcmcn  ajanbmbb(k-1)cmcn = a(j+n)b(m+k)c(m+n) = w
 L1 (k > 0)  L(G)

 L1  L(G)
 L = L1  L2  L(G)
 L = L(G)

Problem 5 (170/3): Convert the following to Chomsky Normal Form:

S aSaA | A
A  abA | b

S  I1I2 | TA | b
A  TA | b
I1  T1S
I2  T1A
T  T1T2
T1  a
T2  b

Method:

( (1) : skipping : G has no -productions )

(2) : Create productions for each terminal and substitute
T1  a, T2  b
(3) : Split productions of length > 2, creating intermediary productions of length 2
S  T1ST1A  { S  I1I2 , I1  T1S, I2T1A }
A  T1T2A  { A  TA, T  T1T2 }
(4) : remove unit productions
SA  { S  TA | b }