IntradayMomentumIndex

Summary

IntradayMomentumIndex is RTTA's streaming Intraday Momentum Index (IMI): an RSI-style oscillator of open-to-close gains versus losses within each bar, aggregated over a rolling window.

Update API

result = rtta.IntradayMomentumIndex(window=14, fillna=True).update(open, close)

Parameter	Default	Meaning
`window`	`14`	Rolling sum window for gains and losses
`fillna`	`True`	If `False`, NaN until the window is full

update(open, close) returns a scalar IMI value in \([0, 100]\) when both rolling sums are defined.

Theory Of Operation

Classic RSI uses close-to-close changes. IMI instead classifies each bar by its intrabar direction:

If \(close > open\), the full open-to-close rise is a gain.
If \(close < open\), the full open-to-close fall is a loss.
Unchanged open/close contributes zero gain and zero loss.

Rolling sums of gains and losses over \(n\) bars form an RSI-like ratio. High IMI means most recent bars closed above their opens (intraday buying pressure); low IMI means the opposite. It is often used on intraday bars or as a day-session sentiment measure on daily OHLC.

Unlike Wilder RSI, this implementation uses simple rolling sums (not Wilder smoothing) of the gain/loss series.

Recurrence

Let \(o_t,c_t\) be open and close, \(n\) the window.

\[G_t = \begin{cases} c_t - o_t & c_t > o_t \\ 0 & \text{otherwise} \end{cases} ,\qquad L_t = \begin{cases} o_t - c_t & c_t < o_t \\ 0 & \text{otherwise} \end{cases}\]

\[IMI_t = 100 \cdot \frac{\sum_{i=0}^{n-1} G_{t-i}}{\sum_{i=0}^{n-1} G_{t-i} + \sum_{i=0}^{n-1} L_{t-i}}\]

When fillna=True, partial windows use the samples accumulated so far. When fillna=False, return NaN until the gain buffer is full. Safe division yields a defined value when both sums are zero.

Implementation Notes

Implemented in src/rtta/indicator.cpp in class IntradayMomentumIndex.
Two RollingBuffers with running gain_sum_ / loss_sum_ via rolling_sum_push.
Output is a scalar double, not a result struct.

Summary

Update API

Theory Of Operation

Recurrence

Implementation Notes

Reference