The TT ADL Analytics block is underway!

ADL Analytics block 1.0 has been enabled on all production Algo Servers. Specifically, the first version supports time bars (historical + live). Documentation can be found here.

As noted above, the first set of time-based analytics is now available in our test environment - https://adl-tt-accept.trade.tt/. We are targeting to release this in our next major deployment. Stay tuned for further details...

Regards,
Andrew

Believe me Daniel, I understand how complex software projects such as this can be. But we're still paying customers (more than 10 years here) and just in the last few weeks I've had to open incident tickets that I am surprised I had to even open for a software that's been many years in the making. Between that, the almost daily bugs I find during trading and non responsive account managers at TT, things are getting a bit annoying now.

I don't want to sidetrack this thread, so will just get back on topic...

Any update on the analytics block? Also, when will it support spreads?

Preview...

I am so excited! LET'S GET THAT DOUGH!

Any picture or video to keep us... excited for the end of the month?!

update on analytics block?

Will it be here for Christmas?

We are targeting to start a beta with a small set of users beginning Dec 15. After we get some traction to shake out any issues, we'll open this up to all users to have a look. Stay tuned for further information. But in the meantime, here's a preview of how it will look and function.

The Analytics block will offer time and volume bars. In this example, we are specifying one minute bars. Notice that the interval is an input so it can be set at start time but cannot be changed afterwards.


The top bar is the one which is currently in progress. The start time ("start_0") for this bar is expressed as seconds since epoch. You can pick and choose which bar outputs that you want to display. When the bar closes, a discrete event is fired from the OnBarComplete output port.

When the algos starts, it will load historical data based on how you've set the parameters, subscribe for live prices, and then stitch these together to form a continuous stream. In this example, we have chosen to store four 1-minute bars with Data Access Method = "ALL". As you can see, this means that the outputs for all bars are displayed as output ports. (Our code name for this internally is called "totem pole mode".) As the top bar completes, the data for the bars will shift down.

You can also choose Data Access Method = "Index". This effectively a data retrieval style similar to the Value Bucket block. A new input port called "Index" becomes visible that drives which bar's values will appear in the bottom set of output ports. The Index value can change at any point.


This example only shows bar data for the Analytics Type. But we will be adding a whole array of analytics which, when selected, will provide their results via additional output ports.

Count me in. Would LOVE to play around with this

Any update on this?