First

Description

First latches the first finite value it sees and returns it from then on. NaN inputs are ignored: they return NaN and leave the latched value unchanged. reset() clears the latch. As a per-bar reducer it yields the opening value of each bar.

Parameters: none.

NaN handling: NaN inputs do not update the latch. The output at a NaN index is NaN and subsequent finite samples continue normally.

NaN handling

Policy: ignore. A NaN in any input at index t causes the function to skip that step: output at t is NaN and internal state is unchanged. Subsequent finite samples are processed as if step t had not occurred.

Examples

Usage example

import numpy as np
import plotly.graph_objects as go
from screamer import First, Resample

rng = np.random.default_rng(42)
n = 100
x = np.cumsum(rng.normal(0.0, 1.0, size=n))
idx = np.arange(n, dtype=np.int64)
opens, bar_idx = Resample(freq=10, agg=First())(x, idx)

fig = go.Figure()
fig.add_trace(go.Scatter(y=x, mode='lines', name='x[t]',
                         line=dict(color='steelblue')))
fig.add_trace(go.Scatter(x=bar_idx, y=opens, mode='markers',
                         name='bar open (First)',
                         marker=dict(color='orange', size=8)))
fig.update_layout(
    title="First: Opening Value of Each Bar",
    xaxis_title="Index",
    yaxis_title="Value",
    margin=dict(l=20, r=20, t=60, b=20),
    legend=dict(orientation="h", yanchor="bottom", y=1.02, xanchor="right", x=1)
)
fig.show()

Implementation Details

First keeps a single double and a boolean flag. On the first finite input the value is latched; subsequent finite inputs return the same latch. NaN inputs return NaN and leave the latch unchanged. Memory is O(1).