Working data
The data you research with, in one place: the symbol, the timeframe, the period and the out-of-sample reserve. With a single 1-minute file you can work on 1 hour, 4 hours or daily, and only on the latest years, without creating any file by hand.
The pill in the main window
The header of the main window has a pill that sums up your working data, for example GBPUSD · 1 h · 2021–2026 · OOS 30 %, with a timeline below it. Click it and the selector opens:
Any of your data files. Its Quote Manager record (point value, commission, slippage) travels with it.
The file's native timeframe or a larger one that can be built cleanly from it: 1-minute data gives 5, 15 and 30 minutes, 1 hour, 4 hours or daily. Those that cannot be built without cheating are greyed out («Cannot be built cleanly from this data»).
All, the last 10, 5, 3 or 2 years, or specific dates. «The last N years» roll forward by themselves: when you add new data, the span moves with it.
The final share of the data (0 to 50 %) kept out of sample. The selector shows the cut-off date and how many bars are left for the search and for the reserve.
If your choice leaves very few bars (fewer than about 300), the selector warns you: with so little data, research is not reliable.
Every window inherits it… or uses its own
The research modules carry the same pill, only smaller. When they open they inherit the main window's data; by clicking the pill you can change it for that window only, without touching the others.
The window is using its own data. «↺ back to the main window's» brings back the inherited data.
What you chose in the module also becomes the working data of the main window.
Changing the main window only affects windows you open afterwards. And while a module is computing, its pill is locked.
What each module does with it
They research with the pill's symbol, timeframe and period, and the OOS reserve replaces the old IS/OOS slider: no search phase sees the reserve, and Performance shades it.
They work with the pill's data. In the Edge Scanner, the in-sample % of the OOS filter comes from the reserve; in the Mentor, the reserve places the IS cut and the slider is still available.
They take the symbol and timeframe, but the period does not cut the data: it places two marks on the date slider, which still sweeps the whole history in real time. The best of both worlds.
The Seasonal Analysis studies only the chosen period. The AQL Editor tests the strategy and sends it to the Portfolio with its pill's data.
Modules that work on an existing strategy (Contrast, 360 Analysis, Multi-TF, Survival, Library, Portfolio) have no pill: the strategy already knows which data it was built with. The Chart and Quotes show the main window's data.
Every strategy stores the working data it was built with, and so does a Strategy Lab template. When you open it in another module, or on another computer, the same data is rebuilt.
For a timeframe or period different from the original, AniQuant prepares a working copy in a hidden folder (C:\AniQuant\Datos\.contextos\). If you update the original file, the copy is rebuilt by itself. Your files are never modified.
You have 1-minute GBPUSD since 2003 (almost nine million bars). You choose 1 h · last 5 years · OOS 30 %: AniQuant prepares about 31,000 hourly bars in under a second, Genesis searches the first 70 % and the last 30 % stays untouched to judge whatever it finds. Tomorrow you try the same on 4 hours by changing a single option.
AniQuant can be tried free for 14 days, with every module and no card. If you have the book, 30.