StrataMEMORY

3-layer pattern memory bank (Short / Mid / Long timeframes). Tracks fake breakout and trend strength patterns to influence bias and trap_risk.

StrataMEMORY()

from strata import StrataMEMORY

memory = StrataMEMORY()

memory.update(inp, state)

Feed current signals and state into memory.

inp   = sense(window)
state = update_state(state, inp, memory.snapshot())
memory.update(inp, state)

memory.snapshot() → dict

Returns the current pattern memory as a dict for use in update_state().

ms = memory.snapshot()
# {
#   "fake_breakout_short": 0.12,
#   "fake_breakout_mid":   0.08,
#   "fake_breakout_long":  0.04,
#   "trend_strength_short": 0.45,
#   "trend_strength_mid":   0.38,
#   "trend_strength_long":  0.31,
# }

memory.reset()

Reset all memory layers to zero.

memory.reset()   # call at start of new trading session

How memory influences state

• fake_breakout patterns raise trap_risk — previous fake breakouts warn of future traps
• trend_strength patterns reinforce momentum — sustained trend behaviour gives momentum persistence
• Memory decays over time (exponential decay per layer)