← Tools

Oracle

Paste a time series. Find hidden frequencies. Project forward. When it can’t predict, it says so — low confidence on unpredictable data is an honest answer.

PASTE NUMBERS (any time series: stock prices, temperature, steps, heart rate, seasonal data)

Loading runtime...

Paste any sequence of numbers and press Run. The oracle finds hidden frequencies and projects them forward.

Everything is free. If this helped → support GUMP

HOW IT WORKS

Spectral decomposition via FFT. Every time series is a sum of oscillations at different frequencies. The oracle extracts the dominant modes, then projects them forward. Confidence decays exponentially with distance from known data — the further ahead, the less certain. Works on any periodic or quasi-periodic signal.

HONEST LIMITS

Cannot predict regime changes, black swan events, or truly random data. If the signal has no periodic structure, the oracle will say so (low amplitude modes, fast confidence decay). The prediction assumes past patterns continue. They might not.

INSTALL

pip install begump
from gump.oracle import predict, extract

freqs = extract(my_data) # find hidden frequencies
future = predict(my_data, steps=20) # project forward
print(future['values']) # predicted values
print(future['confidence']) # confidence at each step

Your data stays on your machine. No cloud. No API keys. Deterministic: same input, same output, every time.

GUMP — Tools · Support · [email protected]