Regarding LOOP - Endloop Performance

[Former Member]

Dear All ,

I am using below code in my prg..

loop at itab.

loop at itab1 where matnr = itab-matnr.

endloop.

endloop.

IN this code itab is having only 1 row for particular matnr and in itab1 is having multiple row for particular matnr. and i am facing here performance issue . it is taking so much time. Because for particular matnr , program have to search matnr in whole itab1. so is there any code line i have to use so it can search directly matnr from itab1 and dont need to search whole table.... because here if itab is having m row and itab1 is having n row , than looping is happened mn times..

so , how can i reduce it ? please anybody know , than please help me for same...

Thank You ,

[Former Member]

Hi,

loop at itab.
loop at itab1 where matnr = itab-matnr.

endloop.
endloop. 

 itab1 is having multiple row for particular matnr

define tab1 table as sorted table with non-unieque key matnr.

and also used the field-symbols instead of work area.

loop at itab Assigning <fs_itab>.
loop at itab1 assigning <fs_itab1> where matnr = <fs_itab>-matnr.

endloop.
endloop. 

Define

field-symboles <fs_itab> has type of itab table type.

<fs_itab1> has type of itab1 table type.

It will give better performance, Don't used the parallel cursor technique.

Try it it will resolve your problem

[uwe_schieferstein]

Hello

You should definitely have a look at the MUST read blog written by Rob Burbank:

[JOINS vs. FOR ALL ENTRIES - Which Performs Better?|https://www.sdn.sap.com/irj/scn/weblogs?blog=/pub/wlg/6050] [original link is broken] [original link is broken] [original link is broken];

Regards

Uwe

[sarang_gujrati2]

hi,

u can use parallel cursor technique to improve performance .

sample code is like :

sort: lt_vbpa by kunnr, "Sorting by key is very important

lt_kna1 by kunnr. "Same key which is used for where condition is used here

loop at lt_vbpa into wa_vbpa.

read lt_kna1 into wa_kna1 " This sets the sy-tabix

with key kunnr = wa_vbpa-kunnr

binary search.

if sy-subrc = 0. "Does not enter the inner loop

v_kna1_index = sy-tabix.

loop at lt_kna1 into wa_kna1 from v_kna1_index. "Avoiding Where clause

if wa_kna1-kunnr <> wa_vbpa-kunnr. "This checks whether to exit out of loop

exit.

endif.

• Your Actual logic within inner loop ******

endloop. "KNA1 Loop

endif.

endloop. " VBPA Loop

Regards,

Sarang

[Former Member]

Hi Sarang

Thanks to give me sample code of parallel cursor technique..

Can u just what is meaning of binary search ? how it works ?

And here in your secong loop

loop at lt_kna1 into wa_kna1 from v_kna1_index.

if lt_konv is having multiples row for particular Kunnr.. than its read only 1 time than how it works ?

[kesavadas_thekkillath]

Can u just what is meaning of binary search ? how it works ?

Please understand binary search,sorted tables etc before getting into the coding

Hi,

You can search the documentations and read how the sorted tables work .

Let me explain how binary search works

consider the below entries in your internal table. You must

sort the internal table before doing a binary search.

You will know the importance of sorting when you understand the functionality.

This is just a rough explanation.Suppose you want to get the record

which holds the value 10, when you use a binary search.

the contents are divided into two segements like

1

2

6

7

***

9

10

13

17

18

It takes the left segement and does a left traversal, if the value is found its

returned else it takes the right segement and again divides it

like

9

10

***

13

17

18

It again does the same left traversal, here the data is found.it stops and exits.

Experts correct me if im wrong

[Former Member]

Hi Keshav,

Thanks for your explanation......

[kesavadas_thekkillath]

Data:itab1 type sorted table of type1 with non-unique key matnr.

Just declare your matnr as above and use the same nested loop.

[Former Member]

Hi ,

Keshav , thanks for reply...in your declaration what is meaning of type1.. And how it works can u explain in brief ?....

[ravi_lanjewar]

Hi,

loop at itab.
loop at itab1 where matnr = itab-matnr.

endloop.
endloop. 

 itab1 is having multiple row for particular matnr

define tab1 table as sorted table with non-unieque key matnr.

and also used the field-symbols instead of work area.

loop at itab Assigning <fs_itab>.
loop at itab1 assigning <fs_itab1> where matnr = <fs_itab>-matnr.

endloop.
endloop. 

Define

field-symboles <fs_itab> has type of itab table type.

<fs_itab1> has type of itab1 table type.

It will give better performance, Don't used the parallel cursor technique.

Try it it will resolve your problem

[Former Member]

Hi Ravishankar,

Thanks for ur help.. Its really works and easy than another method.......