1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
//use crate::common::validate_inputs;
use crate::common_simd::assets::validate_inputs;
use crate::indicators::bbands::{
min_data, output_length, validate_options, IndicatorState, State, INPUTS_WIDTH, OPTIONS_WIDTH,
};
use crate::indicators::simd_indicators::road_train::{Asset, Driver, PrimeMover};
use crate::indicators::simd_indicators::{bbands_simd::calc_simd, stddev_simd::SimdState};
use crate::types::IndicatorError;
use std::simd::Simd;
/*pub use crate::indicators::simd::{
bbands_simd::calc_simd,
stddev_simd::{calc_simd as calc_stddev_simd, SimdState},
};*/
/// SIMD driver that advances the Bollinger Bands (BBANDS) across `N` asset lanes per scheduling
/// epoch.
struct BbandsDriver {
/// Pre-computed `1.0 / period` SMA multiplier broadcast to all lanes.
multiplier: f64,
/// The rolling window size (number of bars summed per average).
period: usize,
/// Number of standard deviations for the upper and lower band offsets.
std_dev: f64,
}
impl Driver<State> for BbandsDriver {
/// Processes one epoch of bars for `N` assets simultaneously using SIMD.
///
/// Reads from `inputs[asset][0]` (real prices), writes `[lower_band, middle_band, upper_band]`
/// to `outputs[asset]`, and updates `states[asset]` in place.
fn next_run<const N: usize>(
&mut self,
inputs: Vec<Vec<&[f64]>>,
mut outputs: Vec<Vec<&mut [f64]>>,
mut states: Vec<&mut State>,
_options: Vec<Option<&()>>,
) {
let len = inputs[0][0].len();
let std_dev = Simd::splat(self.std_dev);
// Optimization 1: Direct array construction instead of collect+try_into
let mut state = SimdState::new(&states);
let multiplier_simd = Simd::splat(self.multiplier);
// Optimization 2: Pre-compute all input and output pointers
let input_ptrs = crate::extract_input_ptrs!(inputs, N, real_ptrs);
let (lower_band_ptr, middle_band_ptr, upper_band_ptr) = crate::extract_output_ptrs!(
outputs,
N,
lower_band_ptr,
middle_band_ptr,
upper_band_ptr
);
// Optimization 3: Simplified main loop with pre-computed offsets
for (j, i) in (self.period..len).enumerate() {
// Get new and old values using pre-computed pointers
let (old_vals, new_vals) = crate::extract_simd_at_indices!(N, input_ptrs,
old_vals @ j,
new_vals @ i
);
let (lower_band, middle_band, upper_band) =
calc_simd(&mut state, std_dev, new_vals, old_vals, multiplier_simd);
crate::write_simd_at_indices!(N, j,
lower_band_ptr => lower_band,
middle_band_ptr => middle_band,
upper_band_ptr => upper_band
);
}
state.write_states(&mut states);
}
}
/// Calculates the Bollinger Bands (BBANDS) for `N` assets simultaneously using SIMD
/// parallelism.
///
/// Bollinger Bands consist of a middle SMA band and upper/lower bands placed a configurable
/// number of standard deviations away. All assets share the same `options`. Uses the
/// [`PrimeMover`] scheduler to batch assets into SIMD-width groups.
///
/// # Arguments
/// * `inputs` - An array of `N` asset input sets; `inputs[i]` is `[&[f64]; INPUTS_WIDTH]`
/// containing the real price series for asset `i`.
/// * `options` - Shared options applied to all `N` assets: `[period, std_dev_multiplier]`.
/// * `_optional_outputs` - Unused; BBANDS has no optional output lines.
///
/// # Returns
/// `Ok((outputs, states))` where `outputs[i]` contains `[lower_band, middle_band, upper_band]`
/// for asset `i` and `states[i]` is the final [`IndicatorState`] for asset `i`.
/// Returns `Err(IndicatorError)` if any input is too short or options are invalid.
pub fn indicator_by_assets<const N: usize>(
inputs: &[&[&[f64]; INPUTS_WIDTH]; N], //stock[ fields [ field [f64] ] ]
options: &[f64; OPTIONS_WIDTH],
_optional_outputs: Option<&[bool]>,
) -> Result<(Vec<Vec<Vec<f64>>>, Vec<IndicatorState>), IndicatorError> {
validate_inputs::<INPUTS_WIDTH>(inputs, min_data(options))?;
validate_options(options)?;
let period = options[0] as usize;
let std_dev = options[1];
let real: [&[f64]; N] = std::array::from_fn(|i| inputs[i][0]);
//init ema, sliced inputs and multipliers
let (simd_state, multiplier) = SimdState::init_state(&real, period);
let states = simd_state.to_states();
let mut road_train = PrimeMover::<N, State>::new();
let mut output_buffers = Vec::with_capacity(N);
for (i, state) in states.into_iter().enumerate() {
let asset_inputs = vec![inputs[i][0]];
let (middle_band, upper_band, lower_band) = {
let capacity = output_length(inputs[i][0].len(), options);
(
crate::uninit_vec!(f64, capacity),
crate::uninit_vec!(f64, capacity),
crate::uninit_vec!(f64, capacity),
)
};
let mut output_buffer = vec![lower_band, middle_band, upper_band];
//let adosc_len = output_buffer[0].len();
let mut asset_outputs = Vec::with_capacity(output_buffer.len());
for j in 0..output_buffer.len() {
unsafe {
//let slice_len = output_buffer.len() - starts[j];
// Get a mutable reference to the output buffer for this asset
let output_buffer = &mut output_buffer[j];
asset_outputs.push(std::slice::from_raw_parts_mut(
output_buffer.as_mut_ptr(), //slice from
output_buffer.len(), // slice to
));
}
}
road_train.add_asset(Asset::new(
asset_inputs,
asset_outputs,
i,
period,
period,
state,
None,
));
output_buffers.push(output_buffer);
}
let mut driver = BbandsDriver {
multiplier,
period,
std_dev,
};
let states_vec = road_train.drive(&mut driver);
let mut states = Vec::with_capacity(N);
for (i, state) in states_vec.into_iter().enumerate() {
states.push(IndicatorState::new(
unsafe { inputs.get_unchecked(i).get_unchecked(0) },
state,
period,
multiplier,
std_dev,
));
}
Ok((output_buffers, states))
}