Binary Search (2)

For Example, the data is:
The key is 2

Record	[1]	[2]	[3]	[4]	[5]
Data	2	5	21	1	6

First step: the data must be sorting. This sample, we can sort the data with bubble sort methods.

Record	[1]	[2]	[3]	[4]	[5]
Data	2	5	21	1	6
Step 1	2	5	21	1	6	2 and 5, no Exchange
	2	5	21	1	6	2 and 21, no Exchange
	1	5	21	2	6	2 and 1, Exchange
Step 2	1	5	21	2	6	5 and 21, no Exchange
	1	2	21	5	6	2 and 5, Exchange
Step 3	1	2	5	21	6	5 and 21, Exchange
Step 4	1	2	5	6	21	21 and 6, Exchange

Second step: we can search data key 2 with binary search

New data:

Record	[1]	[2]	[3]	[4]	[5]
Data	1	2	5	6	21

Step [1]:

                (low+high) div 2 =  (1+5) div 2

                                            =  6 div 2

                                            =  3

                the data on the record [3] is 5, then 5>3, next step swipe left

	low		middle		high
Record	[1]	[2]	[3]	[4]	[5]
Data	1	2	5	6	21

Step [2]:   new_high = middle-1 =3 - 1 = 2
                (low + new_high) div 2 = (1+2) div 2
                                                    = 3 div 2
                                                    = 1
                the data on the record [1] is 1, then 1<2, next step swipe right

	low	new_high
Record	[1]	[2]	[3]	[4]	[5]
Data	1	2	5	6	21

Step [3]:  new_low = low + 1 = 2
                (new_low + new_high) div 2 = (2+2) div 2
                                                    = 4 div 2
                                                    = 2

	new_low=new_high
Record	[1]	[2]	[3]	[4]	[5]
Data	1	2	5	6	21

                the data on the record [2] is 2, end.

REseArCH 1

SISTEM PERHITUNGAN LAMA PENYINARAN MATAHARI DENGAN METODE OTSU THRESHOLD (STUDI KASUS: ST. KLIMATOLOGI BARONGAN)

Asih Pujiastuti, Agus Harjoko

Abstract

Solar radiation measurements in climatology station barongan done by using campbell stokes and record card. Solar radiation is the accumulation of the whole length of the object radiation recorded on the record card. Solar radiation data reading performed by observers by estimating the long of object irradiation compared with the scale of hours listed on the record card. The difficulties encountered by observers in conducting solar radiation data readout is the shape of the object radiation recorded on the record card are not always in the form of elongated areas, some areas disjointed, even can only form small holes along the track on the card record. The purpose of this study was to design and build a system that can be used to calculate solar radiation recorded on the card record, thus increasing the accuracy of measurement data readout. Cropping process is carried out in order to obtain the right area for analysis. The approach method in object segmentation is done by implementing a solar radiation threshold otsu method. The test results showed that the determination of the radius of the opening and removing noise value appropriately irradiation can obtain the number of objects on test images with RMSEof1.4429 compared with the amount of exposure o f the image of the original object. Results of calculation solar radiation by the system showed system RMSE o f 0,51 to observer calculation.

http://ejournals.stta.ac.id/index.php/compiler/article/view/166

Binary Search (1)

For Example, the data is:
The key is 6

Record	[1]	[2]	[3]	[4]	[5]
Data	2	5	21	1	6

First step: the data must be sorting. This sample, we can sort the data with bubble sort methods.

Record	[1]	[2]	[3]	[4]	[5]
Data	2	5	21	1	6
Step 1	2	5	21	1	6	2 and 5, no Exchange
	2	5	21	1	6	2 and 21, no Exchange
	1	5	21	2	6	2 and 1, Exchange
Step 2	1	5	21	2	6	5 and 21, no Exchange
	1	2	21	5	6	2 and 5, Exchange
Step 3	1	2	5	21	6	5 and 21, Exchange
Step 4	1	2	5	6	21	21 and 6, Exchange

Second step: we can search data key 6 with binary search

New data:

Record	[1]	[2]	[3]	[4]	[5]
Data	1	2	5	6	21

Step [1]:

                (low+high) div 2 =  (1+5) div 2

                                            =  6 div 2

                                            =  3---> middle

                the data on the record [3] is 5, then 5<6, next step swipe right

	low		middle		high
Record	[1]	[2]	[3]	[4]	[5]
Data	1	2	5	6	21

Step [2]:  new_low = middle +1 = 3 +1 = 4

				new_low	high
Record	[1]	[2]	[3]	[4]	[5]
Data	1	2	5	6	21

                 

               (new_low + high) = (4+5) div 2

                                             = 9 div 2

                                             = 4

                the data on the record [4] is 6, then 6=6 , end.