Usually the indicators do not interfere. This means that the CPU load caused by a specific indicator is the same, whether it is used alone or in combination with other indicators. Of course several badly coded indicators can add up to exceed the threshold.
There are a few things that you can do to reduce CPU load
- first use all your indicator with CalculateOnBarClose = true
- second do not change the display updated interval to a lower value than 0.5 sec
You may also post the source code of the indicators - or in case that you have downloaded them from this forum, post a link to the source file - and we should be able to tell you which of the indicators cause the CPU load. A general statement does not help. We need to go into details.
You can further conduct individual tests for each indicator to determine the CPU load as follows:
-> download Microsoft Process Explorer
-> download one week of replay data for ES 12-15 (it is free)
-> open a virgin chart for ES 12-15
-> add the indicator to the chart and set it to CalculateOnBarClose = false (maximum damage)
-> start market replay at maximum speed (x 500)
-> open Process Explorer and watch the CPU load for all kernels of your processor
When played at speed x 500, the regular session for ES should take about 50 seconds. If it takes a few minutes to complete you know that you will be running into problems.
The following 9 users say Thank You to Fat Tails for this post:
I to use 4 screens and look at a lot of data, NT7 is designed to stuff all processing through 1 core and is very laggy. There are a few things I have learnt that offset this and just about make NT7 well half usable...roll on NT8.....or is that just a myth
I got this info from the Jigsaw site where you will find in detail how to do it all in the forum (Petes a saint):
You have to minimise disk writes so as not to bombard disk with unimportant stuff.
1 - If you dont need market reply data, turn it off
2 - Disable search indexing (youtube video)
3 - Download Process hacker:
Go to Processes tab
A - Right click Ninja, Affinity, deselect, select 1 core (I the 2nd)
B - Right click Ninja, Priority, High
C - Right click Ninja, I/O Priority, High
Once you have done this, get task manager up and you will be able to see Ninja raping the chosen core and you will be giving it the best chance to not let you down by clogging your disk up.
You will have to do this every time you start Ninja, but it only takes 30 seconds. Hope this helps
Last edited by traderadam; October 31st, 2015 at 06:47 AM.
The following 4 users say Thank You to traderadam for this post:
Thanks FAT TAILS for the insight. I have narrowed it down to 3 indicators I've imported from the forum. Each causes chart lag on their own. I've changed system parameters as outlined and still have chart lag. I will most likely need to hire someone to have these three indicators recoded. The VolStop indicator has a known issue of not being able to save default colors.
Favorite Futures: Gameplay Klownbine® Trading of Globex
Posts: 1,276 since Jul 2009
Thanks: 1,227 given,
The BMT wave indicator was fixed in 2010 (by me), so nothing to be done there. The swing indicator, if running COBC false, does not have anything to prevent processing of redundant intrabar ticks (same price as preceding tick). It's a total waste to process redundant intrabar updates. Generally at least 70% of intrabar ticks are redundant. Best bet would be to make sure swing indicator is running on COBC true.
Be careful about setting Ninjatrader priority to high, can bring your PC to a standstill.
Your strategy will probably work MUCH better in NT8 with its multi core multi threading.
Don't forget about the Ninjatrader setup hints in Post #7.
"If we don't loosen up some money, this sucker is going down." -GW Bush, 2008
“Lack of proof that something is true does not prove that it is not true - when you want to believe.” -Humpty Dumpty, 2014
“The greatest shortcoming of the human race is our inability to understand the exponential function.” Prof. Albert Bartlett
The following 2 users say Thank You to Zondor for this post: