Oncga.
Theta
( B
Big
Notationg Asymptatie *
ethod countFyquany
fhuniomplexit,
ngtire Tie
Arelysia
Corect/EMcient
iDeqinite
lfinite
DOutput
picblenm patkular
to
sohve instctiong finite
ostaps *Alaoithm
Mondoy
Paga
Date.
clasSMate lab-l0"3
leost
i leost
:0;i<n-1it) forinti
(T)[Link]-
let
0 a: ey
isÇ,icnit+) torint
solt Bubble ’
’Selection
Soft
sortTnserion ’
techriques Jotting *
Page
Date
classate
classMAie
Date
Page
Ssuap(a(loos]eda])
h-|130 alssmate
Date
Tuasday Page
there exist,cng'o han
then we (an
Say i
olgl)
2
fn)< c"n2 nn, nl, Vopo boun
3,3 n 21 n
auntio
2
1<lbq n <n < naglogn<n<n<
So,we
4)2c"n
ntn*1
1
2 n
ereatd tig etenteA times Noo
fonst onst
end
end
RetSurn
end
n-l to
ko rtor
nto tori
1
n-1 toio for
ntp,la2,) (A, Sun ()
Pqe
Datc
T prod> 1
for (i1ica iia)
prod prod *i
Toke log, on both jides
k2 logn
(ongt.
nsertionsort (int A[], int n) timey crentd
for (nt ie1;i<nié)
C
hile
A]A];
n-1
Th-10:30 classmate
Date
fsiday, Page
Renning tinse d Insertion sort.
+k)+4(n)
Woist 'cOse
When array u Yeveise Soited
Jaleas L 1 2 h
toop ereats
Tan
Sbatuking in (
+)n(6+4tGt-g-4)
<4-y-c)RTO(0
classmate
Date
Paqe
Best cose
Alveady sozted.
"L1 n1
t.1al1
So
Substituting in
+s[nt)
RT =O
Average eCaze
Sorne
t i,or exary value
So t
-1 |=1
On your Own )
clhssrute
Date
Page
k Loop Tnvaient
Helps ta provo' (orectness o an
olgaátha
Set ostatements hich are tue befoe after eoch
loop itetation .
3 3steps ) Initiation
i) Maintanante
)Tetmindtioa.
(ovectness o Insertion sot
Staten. Alo-j-2u sotted bejos h teration of loop
obviously
.Initialisation; Alo] toially sotted
" Maintenane i ih iterathon we
we wilL
will move elements
A-aA(o] by
wntill (orvect position o
1 position to right
element u found
So, ofterlaop terminathion A[ousorked
Tetmination nt1
Aray A[o-n-1 will be soted.
*Seection Soit - leost no-o swaps (h'swaps)
Code Void selection sort (Int arr|,int n)
onstTimes
cKecuted
1
tnt i leost
tor (iao;i<n-1;it+)
leost = i C3
for(j ait1;j<n;jt)
arY
4ar i) < leost])
leost
n-1
4iilleost)
saop(ar 1, a [least]): n-1
loop
n-lJ
"RT-O(R) for best,
worst avg cose8L
"n-4
n-1
n-2 n- 1]1
Solting
Intenal Extemol
Solting softing
Eschonge Comparson Met g
sot t
e Merge
Sort
Selection + Bubble
Insertion
LHoap + Shell
bQuick
*Bubble Sott - Bubble snk.
E 3,5,65,3,3,12
Bss1; 634, 75, 65,3,2,13,12
6, 34, 65, 45 3,2)13,12
6, 34, 65 3, 3a, a, 13, 1à
6, 34, 65, 9,5, 3, 12
6346$ 3;a3, 5, 13
6,34,65 3, R3,2,75
ses. al o() nln=) omparisong* Total
uhere time Running
-1) ),an (a
Suop
;j<ij) n-1 jz(int tor
Vp: Bubble
t, Smalest i=0;icn-1,
it+) (int far
Vesion ) an],
int bubblesost
(int vOid :
Code
1312, 3,
65, 75, 6, 34, 2,
12,13 65,3, 6,6, 2,34,
13
1R,34,6R,?S,
65,3
3,12,
13 34,6,5,
65, 2,
13 ,3,
1R, 6,ts65, 34
565,3,23 6,34, 1:Poss
3,2,? 34,6,95,s5,
upBbble
Pag?
Date
lssMate
Face
Void bubblesort (int ar[]intn ’Version )
for (int i0isn1; it+) Lavçest element,
Sink Down.
tau int jojca-i-1j*)
4(ar]> am(jil)
Jmploved
(ode void bubblesort (int arl, intn
bool flaq True:
o(intiáo;i<n-1 flaqit)
I ar i <an (i-1)
Suep (an lj,a 1)
flag kue
Ruaning Time here u
akead sotbed( best case)
Date
Pan
Sott -Batti sizeo dota u more thon
4, 13,6,4, 16, 3,12
4,1362 I4, 16,3 la
4,13 6,2 1,15 »3,12
|4,13| 4,15 3,12
2,4,6, 13| 3,12,14,15
hl :30 chssMAte
Date
Tuasday Page
* Merge Soct:
Bozed on Diide & Conquet
Code Noid
Naud merge_solt (int A,at bia iot end)
mid =
All 4 lins
are inside the
mergesoit (int A, int bi,int mi)
i loop.
merge -sofA (int Aiat midr1 int end) S
merge_ Seork (int A,int beg iat mid int end)9
ige A, bg mid,gnd)
nQend -mid
tor opying value romn K
L) A[Leq +i-1aioy to toiay
foj:13<najt) Har
R.
coping value fom k to
L[n1+1 use nt max instead
R[hat1j 0
elassmate
Data
Page
else
* Complexity f Meige Sodh i
’ Master TheoYem -
as subproblenm
aTl)+{6)
subproblem
" ReLurrence M.S
n1
1 n* 1
So
ASE KASE2Í ) CASE
’
4 Kecuesion :14 3 1:
Then, Now
So, a22&bd if
f)
f)o(log y fn)<
ree :no
log,
nogthen polynomially
ti
then
4Cn n thTlena) e(
n-
T( Smaller
n)
)
(nlogn) e(()
Defu
Pcçc.
casamte
Bucket
Shell
CountingRadix
X Quick
Heap Lept
X
Merge
Bubble
Selectien
Tplacs 'Sabiliy Thserton i- Done
So,T(
TFreResursion
e Heioght XCn
R-T=
log, i nlog
log take
sides both on
Page
Dat:.
classate
Daic.
Page
aesday
Sost
Heap
HeigktHeap
Haigthto node y u longest pat ftom 'noda Sy' to ha
Defi
leof (aunt the edges)
Suppoxe 'H' binary hesp oi n elemets then heigkt
log n 2 h lo
Analysis o MAK- HEAPEY
Maxheapity tues the irclathiorship bl roct. node 4 ibs
chidren
RT= 2x Tree height
Sub ) in )
"RT lag,n (o) olgh)
AE each lavel, conmparisons aremade, for nade i' to
csSMte
Date.
Page
32 \6 75 16
Left Void heapsort(à)
budmarheap (A)
o A-lengih doun to Q
do exchange Ala A[1
A: heapsize Aheapsize-1
mak-heapity(A,1)
48)
16
Code:void budmat heapA)
A heapsize = Ailength
for i s A-length down to 1
masheapty (A,)
3a
Stu
cssnate
Daic_
Paco.
48.
15
48
45 16 19 G
32
-Einal tax Heap
max henpit (4, )
Righk (i1 feturn'a rehum atit1
larsest
else
4
largcst
pmas_beapif (A, largost)
h-l330
uesdoy
ooz|24
* Heap Sort Analysis
) ma
-heapty ()
This unctan traes pat from root to leat
So, RT o max heopih x heght o tree
32X log,n
Ranunene velahion af the RT of más heopity)u
descibed as i
T)Ta)+ o()-0
an a ubtrea sizei ooted at
a given node (;'i
the time to
n the may-heapfy functan.
Worgt CO8e scenano g mar-heapiy happens when lost
clhssmate
Dale_
Page
Yeruience() sin moster's theoren
Sobing
a 1
Sioe n(oe applhes
So T) n lagzn)
Build-mahecp)
dn Ih buildarheap) wen
sthe RTdeperos oni howforian max-henpty)
i
alled,
élement miabt shif
down before the plaress tefominotes. Atthe ost leel
there are nodes we do not call mox
tor thenmAt the a most botom leve there are a
ngde e each node might shitt doun to the bettom level.
inode may shit down by level:So
h
T=Ëa
akingihas
classmate
Date_
Page
he may no -ot nodes n in a heap u given byi
ht
Sub () in )
TOn+1
Cach on- cally to
Mayhenpit talles O(log,n) So, Heap oit tokes i
*Shous thaony subtee o nmax heap,the xoot
PyQ-4 (ontaing the larqest valuei
Ang Assume the claly u folze a subtee whose voot 'y not
the lorgest eleman)
Then mar elenment u plesent at more than 1 locaton
et be theinde where may eleent
Since mayx clement u not at soot o! the subtee. So the
hode en' hos a parent such that Alpatent (n< A|n]
But he
may
may
heop opety we must hae Aipornt)>Ap
Dcde,
Page
Hence, our assumpton u fale the statement Hhat cot
contins the lotgest value te
’ Min n'o elerments may.x-haap
N
min
loct, loul hor only 1 denent
element.
’ Max no. Gt elements in mar heap
N
Q A d- no avay made in a max heap uherc a voot hos
d childtern,th¿n iite tha equations or the paventylept
ight node.
Q. Height of d aray heap qven in elments in it
Date
* Quck Sat
Baved on duide & conquer shrategt
Alpi ARjwnseen R
Iitiely ’ o i lost element(n here)
coses
Aay CA à dividud in 3 ports
unseen elenents
clessite
Date
PLge
146
Afpq]ápit
odei Void
potition {A,p,n)
Quickot (4,41,1)
yojd parition (A,p)
to
i=+1
Sasp(AA)
Yetun Í+1
Best Avg olnloga)
Worst
Cache optimizaiarn
alassmate
Date
Page.
* RT Quict Sort t
. Q:S u veausive alao. in'which best uoist /av (ases
are calalated bosed on the selechon o the pivot elemant
a medium yole af ariay
u Smollest /larqestHhen
7
woLt toe will ocur
Recutrence equation
piot
Best Cae -
ODD:2 subarways o size n"1 will be ceated
We can say,siza ahdst nla
n EVEN: 1 subárvay o size na anothes n-1.
Atroost n]a
)So, T() T (n (aplying mout's tcoen
a=2b2,)=n n n toga
Cose d agplie where ln h
Phone No.
Dote
KO Reusion Tree.
Best cose
na
1 1
1 lag nJibg2
=
RTnX Height oi tee nloq
’ Wotst Case; This occuts whain
when pivot iuWsmallagt/latgat elennent
To)Tn-)+ (en
Replocing 'n by n-i)
Tn-)+C-n-)-l
Sub (3) in
Tl-1(a-)Cln)+
Putting n-2 in
)
Con- )
T(a)7(n-3)+
Sub4)in 9)
C-(n-) -)
T() 2T(n-)+ c-[(a-+ (n-)an
chssmc
Date
Pac2
finally i
Student' Sigr
h:10.30
* lounting Sort i- Stable sosting tacdnigue
Sortng in Linear time
where Yonge o valuesOk
oAlgoithm
bomtsart (AB,)
for =1 to A-length
A
no îontains no o times clemgnt i occutsi in atfay
tor i*1 to k
now (ontains ao o elenments des than or equal to
JEAlenh dewn to:1
aCLAL) A1
cALIca)]-1
clasSMate
Date
Page
Dranock
À loO elements
k C
Gunting sost à not feosiba whun 'K' ia very lorge The
uppr bound on kuoCk C-nlc:2
olnt)
Q t doent work on (-ve) vales becau nos ate wed
o indexes in Caray
3: Cannot be applied wwhen yaey are distnet and have
y bigh d blavelws Eq'651,1, A999)
Louer Bound or Comparison Soting Techiqu
lower
prove
Any tomporison sott muet make
in themost coses. to sot n' elenents.
Datc
Ferc
Sostog algo. in tora o decsion tee
Yes No
akb
No
tes ascNo Yes
<
bee -cKasb bcacc
axc<b C<b<a
The d-tree wors Corvecty in al) the
possibk permutation appeoa a leo nole in
he tree
The langh o4 the longat path tom the toot ot the
decision tree to any o its Yeoch oble leot shos tha
wost coe no:o comparisons that any coparison saing
tacbaique con purtoi
Thus, lower bond on the hat a the tree in which
cach permtation appeor a Seach able leof u the louer
bound Yanming tine o Conpaison oting trec.
a d-tee hoa hat leoves for 'a' elernt
Since theye. afe petmutaion) o inputt sequene hat
ppeors os leaf. Sa,
alss
Sine binany tree o hgt 'h' hor nor mofe thon
leaves so
lonbining
USicg Stalinqs apptrximatisa.
log,m!nlogag)
So,h 2n log,
h2nloq n
h3ln loqn