COT5405 Analysis of Algorithms
Homework 1 Solutions
Four questions were graded, 25 points each.
Solutions
1. (a) Similar to binary search, takes
O
(
lg n
).
SEARCH
(
S, N
)
r
= 1
, s
=
N
while r
≤
s
do t
=
floor
((
r
+
s
)
/
2)
if t
≤
S
[
t
]
then s
=
t
else r
=
t
+ 1
if r
==
S
[
r
]
then return r
else return
0
(b) Takes
θ
(
N
)
MaxSubSum
(
S, N
)
lMax
= 0
, gMax
= 0
for i
= 1
to n
do lMax
=
max
(
S
[
i
]
, lMax
+
A
[
i
])
lMax is maximum of all sequences ending with
S
[
i
].
gMax
=
max
(
lMax, gMax
)
gMax is maximum of all sequences ending in
S
[1
..i
].
return gMax
Grading:
Not graded.
1
This preview has intentionally blurred sections. Sign up to view the full version.
View Full Document2. (a) Pass array by copying all its elements.
T
(
m
) = 2
T
(
m/
2) +
θ
(
n
) +
θ
(
m
)
but
m
≤
n
.
T
(
m
) = 2
T
(
m/
2) +
c θ
(
n
)
T
(
m
) =
θ
(
m n
)
OR
T
(
n
) =
θ
(
n
2
)
(b) Copy elements of required segment only.
This does not change the original recurrence of Mergesort. So it takes
θ
(
n
log
n
). (Your solution should verify this.)
Grading:
First part 15, second part 10 points.

10 points for omitting the
θ
(
n
) part in the recurrence and getting
θ
(
n
log
n
).
3. Store
n
intervals in a data structure which holds right endpoints and
left endpoints of intervals. Sort the data structure according to left
endpoints of intervals using any of
O
(
n
lg
n
) sorting algorithms (Merge
sort, Quicksort, Heapsort, etc.).
Let,
I
(
i
) represent the i’th interval data after sorting phase,
a
i
and
b
i
be left and right endpoints of the corresponding interval. Let,
T
be
the total length of intervals and initially
T
= 0.
Scanning all the data structure starting from
i
= 1 to
i
=
n
we need
to consider 3 cases:
•
Case1(
b
i
≤
a
i
+1
):
I
(
i
) not intersecting with
I
(
i
+ 1) then
T
=
T
+ (
b
i

a
i
)
•
Case2(
b
i
> a
i
+1
AND
b
i
≥
b
i
+1
):
I
(
i
+ 1) is included in
I
(
i
) then
T
=
T
•
Case3(
b
i
> a
i
+1
AND
b
i
< b
i
+1
):
I
(
i
) is intersecting with
I
(
i
+1)
then
T
=
T
+ (
b
i
+1

b
i
)
This algorithm correctly computes the total length of n intervals given
This is the end of the preview.
Sign up
to
access the rest of the document.
 Fall '08
 UNGOR
 Algorithms, Binary Search, Big O notation

Click to edit the document details