feat: implement tide chart dashboard (fixes #4)

[R-Panic]

Summary

Implemented the Tide Chart Dashboard to resolve Issue #4.

Tide Chart is a Streamlit view parsing the probability distributions across 5 asset classes to expose their relative strengths compared to the base $SPY.

Resolves #4

Changes

• Built data_engine.py parsing SynthClient forecast vs volatility outputs.
• Developed main.py using streamlit + plotly to render median vs 5th/95th percentile skew cones.
• Highlighted strong Relative-to-SPY movements logically via a reactive Pandas DataFrame Rank Table.
• Implemented pytest suite testing data mapping behaviors.

[e35ventura]

please add Submission Requirements

1. Demo video — Screen recording showing the dashboard in action

2. 1-page technical explanation. Should cover what the tool does, how it works, and how it uses Synth data

[e35ventura]

this needs to look way better to be considered

[R-Panic]

Here are the requested Submission Requirements for the Tide Chart:

1. Demo Video

---

Tide Chart Dashboard: Technical Explanation

What the Tool Does

The Tide Chart Dashboard is an analytical visualizer built on top of the Synth API. It provides a comparative layout of expected asset performance over a 24-hour horizon, focusing specifically on equities: SPY, NVDA, TSLA, AAPL, and GOOGL.

Rather than viewing these assets in isolation, the Tide Chart normalizes their probabilistic price paths to highlight relative strength and weakness against the base market index (SPY). It equips traders with a clear view of which assets possess the highest upside deviation (skew) and the tightest forecasted volatility, allowing for optimized pairs trading or directional bet-sizing.

How It Works

The dashboard is constructed using a two-part architecture:

1. The Data Engine (data_engine.py): A backend processing module that orchestrates calls to the SynthClient. It aggregates raw percentile forecasts and translates absolute price projections into normalized percentage changes.
2. The Streamlit Frontend (main.py): A reactive user interface utilizing plotly.graph_objects and pandas. It visualizes the processed data through two primary components:
   • The Comparison View: A box-plot representation mapping the 5th percentile (downside tail), 50th percentile (median), and 95th percentile (upside tail) expected moves.
   • The Rank Table: A live-sorted dataframe highlighting directional skew and relative-to-SPY outperformance.

How it Uses Synth Data

The tool leverages two specific endpoints from the Synth API (Subnet 50), which serves 1,000 simulated price paths generated by decentralized AI models:

1. Prediction Percentiles (get_prediction_percentiles)
Instead of simple point forecasts, the engine extracts the 24-hour ensemble distribution. It isolates the 0.05 (worst case), 0.5 (median expected), and 0.95 (best case) percentiles for each of the 5 equities. These percentiles are then geometrically normalized against the current price to calculate uniform percentage moves.

2. Volatility Analysis (get_volatility)
The engine fetches the average_volatility array over the same 24-hour horizon. This metric is injected directly into the Rank Table, providing context on whether a forecasted median move is statistically significant given the expected market turbulence.

Derived Synthetics (Skew and Relative Strength)
Using the Synth percentile data, the Tide Chart computes custom synthetics:

• Directional Skew: (95th % - Median) - (Median - 5th %) to reveal if the decentralized ML models are modeling a fatter tail to the upside or downside.
• Relative Strength: Asset Median % - SPY Median % to instantly flag if a stock is predicted to outpace the S&P 500 benchmark.