Build crypto data tables and custom formulas

Tables are live research workbooks

A table combines symbols in rows with market, indicator and formula columns. Sort, filter and inspect the same calculation across a universe without opening charts one by one. Saved tables preserve the research structure, not a frozen screenshot.

• Market columns for price, change, range and volume.
• Indicator columns with configurable parameters and plots.
• Formula columns derived from other series.
• Comparison matrices across symbols and timeframes.
• CSV-oriented copy and paste for external review.

Formula Lab turns an idea into an explicit program

Formula programs run in a deterministic vector engine rather than JavaScript. They cannot access the browser, network or filesystem. Every historical reference is causal: close[1] means the prior closed value, never a future candle.

1. 01

   Name the concept

   State what the series is intended to measure, such as volatility-adjusted momentum.

2. 02

   Choose inputs

   Use OHLCV, quote volume, trade count, candle anatomy, taker delta, CVD or supported injected derivatives fields.

3. 03

   Write the program

   Define sequential variables, use close[1] history, condition ? yes : no branches and the technical, statistics, event and risk function families.

4. 04

   Plot and inspect

   Place a research result on price or in its own pane and inspect missing or unstable periods.

5. 05

   Reuse it

   Add the saved formula to Tables, Screener ranking, Screener conditions, custom strategies or agent research.

Code a complete strategy in the same language

Switch the output type to Strategy position. The final series must represent the desired state: +1 long, 0 flat or −1 short. signal(long, short) encodes one-bar events; position(long_entry, long_exit, short_entry, short_exit) carries the position until an explicit exit or reversal. Formula Lab immediately runs the result through next-bar-open fills with fees and slippage.

Normalize before comparing unlike markets

Raw volume, price change and volatility are not directly comparable across BTC and small-cap markets. Use returns, percentages, z-scores or rolling normalization where appropriate, and keep the normalization window consistent with the decision horizon.

Validate every derived series

• Check the first valid bar and warm-up period.
• Inspect division-by-zero and missing-data behavior.
• Confirm no future bar enters the calculation.
• Compare a few values with an independent hand calculation.
• Test stability across symbols, timeframes and market regimes.