Home >> Support Board

Support Board

Date/Time: Fri, 26 Apr 2024 05:03:18 +0000

How to Post a Message for Support

Post From: New Offering - Order Flow Trading Bots aka SceetoBots

[2018-09-30 13:57:21] Link To Thread - Link To Post [Go To First Post]
TapeReader - Posts: 88
Thanks for putting this out there as it has finally forced me to write some background on why there is a memory hit, and some Best Practices/Lesson's Learned on how to configure the platform to be more efficient with memory. There is a large memory allocation with SceetoBots. The 'Not Responding' that you are experiencing will disappear once the initial launch/load of the studies settles. Some thoughts on why SceetoBots use so much memory: SceetBots is has an R&D component, a order flow processing component,a SceetoBot configuration tool, and self-contained trading orbs. When you load a full chart of all of this stuff, there are not only a lot of core components that are readied for memory, all of the data that any trading orb might need is also loaded into memory a separate time in a separate memory space so that the trading orbs execution responsibilities are never hindered by another data request or CPU request when it's time to get into a trade. This benefit of the decision to replicate data so that it could be used exclusively by the trading orbs, and not accessible by your 'visual'charts, pays off when your Sceeto's trading rules allows it to get into a trade. Our platform is tuned to get your orders into the book at GLOBEX (or other exchanges) as soon as possible. Additional memory allocation for this is a good use of memory. Hence, we decided to not be too frugal with RAM, as speed on the limit entry order can provide more upside than managing high RAM downside. So, your experience of this memory hit is because with the templates that we provide, all of these tools are loaded simultaneously - and this does come with a memory hit. However, there are a few configuration tips that you can use to make your experience lighter. Some thoughts on how configure, and use, SceetoBots as efficiently as possible. 1. Aggregating the responsibilities of visualizing data on charts in one chartbook while keeping all of the Sceeto's trading responsibilities into another chartbook. 2. The chartbook where you are running your Sceeto's for trading should only contain 1 or 2 days worth of data. There is no need to load 50 days of a bunch of deeply processed order flow data into a Sceeto's memory, when the Sceeto's trading responsibilities are focused on real-time data. 3. If you are not doing R&D, do not load large amounts of days on your charts. 4. If you are doing R&D and have 50 - 100 days on your charts, you will have to wait for the memory settle. However, there is a tremendous upside to waiting the few minutes to load your chart. Sceeto ships with a reporting tool that takes advantage of this large memory allocation...and makes doing back-testing research unbelievably fast. Here's a video on it's functionality. https://www.youtube.com/watch?v=hVchxHc50Ic&t=143s What is great about this research tool, is that it allows you click, click, click, to reconfigure a Sceeto, and then click the 'Dash' button and within a second or two, your back-tested results will be presented to you....which is a much nicer experience of having to run a 5 minute back-test every time you want to test one configuration change (This is where a lot of the memory hit goes - but its a good use of memory). Apologies for this initial on-board confusion that you experienced. We are working hard on making the 'getting up to speed' experience better, which includes the following functionality: 1. Change Sceeto study menu option that is exposed when right-clicking on a chart per graphic 2. Change study name from SceetoBots to Sceeto 3. Change the directory location for the sceetobots_config.xml file from the SceetoBots folder to the SC\data folder \\ (This will allow for a much easier experience to share/swap/save trading orb configurations). 4. Update default settings in Sceeto study for less initial configuration by the user. 5. Add input to Sceeto study that automatically launches the Global Launch Pad - where the default is to launch the Global Launch Pad. \\ Super-Important 6. Add input to Sceeto study that allows the user to automatically launch the local chart Launch Pad - where the default is to not launch the local chart Launch Pad. We have just completed development of a new build of SceetoBots that is scoped to make the initial launch experience better. This build is being tested and will be rolled out by Wednesday. On another note, I think this new method for trading is going to be an astounding site to see....the Sceeto's have entirely upgraded my trading experience, and my lifestyle. Beyond the software, we have developed a workflow for using the Sceetos that took a while to evolved. We have been hosting workshops on this new man & machine trading workflow...and have some documentation on it being groomed. One final thought that I think is important to communicate, when you are loading SceetoBots, you are loading your own entire quant trading firm onto your computer...so though, it is a lot of stuff...it's great stuff to have at your fingertips. :-) If you ever get stuck, please reach out to us either via this Sierra Chart forum or via our help desk which you can access here: https://sceeto.zendesk.com/hc/en-us/requests/new We appreciate the effort that early adopters have put into this experience and greatly value your comments.