RTTADocumentation Built by Goblin Reactor ↗

DeMarker

Incremental, causal technical analysis documentation

Summary

DeMarker is RTTA's streaming DeMarker oscillator. It compares rolling upward high extensions to downward low extensions and maps them into a \(0\)–\(1\) ratio.

Update API

value = rtta.DeMarker(window=14, fillna=True).update(high, low)

The first bar contributes zero DeMax/DeMin (no previous high/low). With fillna=False, output is NaN until the window buffer is full.

Theory Of Operation

Tom DeMark's DeMarker measures demand pressure via positive high-to-high changes and supply pressure via negative low-to-low changes. Averaging each over a window and taking the ratio yields an oscillator bounded in \([0, 1]\). Readings near \(1\) indicate persistent upward high extensions; near \(0\) indicates persistent downward low extensions. Extreme zones (classically near \(0.7\) / \(0.3\)) are used as overbought/oversold references.

Recurrence

Let \(H_t, L_t\) be high and low and \(n\) be window (default \(14\)).

\[DeMax_t = \max(H_t - H_{t-1}, 0), \qquad DeMin_t = \max(L_{t-1} - L_t, 0)\]

(with \(DeMax_0 = DeMin_0 = 0\)). Maintain rolling sums over the last \(\min(t+1, n)\) samples:

\[M_t = \sum_{i \in W_t} DeMax_i, \qquad N_t = \sum_{i \in W_t} DeMin_i\]
\[DeM_t = \frac{M_t}{M_t + N_t} \quad\text{(safe divide)}\]

Implementation Notes

The recurrence is implemented in src/rtta/indicator.cpp in class DeMarker using two rolling sum buffers (demax_, demin_).

Reference