chapaty 1.3.0

An event-driven Rust engine for building and evaluating quantitative trading agents. Features a Gym-style API for algorithmic backtesting and reinforcement learning.
Documentation 
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
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298

use std::collections::VecDeque;

use serde::{Deserialize, Serialize};

use crate::indicator::{
    config::{EmaWindow, SmaWindow},
    streaming::StreamingIndicator,
};

// ================================================================================================
// SHARED: Exponential Weighted Mean (Base Logic)
// ================================================================================================

/// Internal helper for EMA-like calculations (Standard EMA and Wilder's
/// Smoothing).
#[derive(Debug, Clone, Copy, Serialize, Deserialize)]
pub(crate) struct StreamingEwm {
    alpha: f64,
    current_mean: Option<f64>,
    window_size: usize,
    count: usize,
}

impl StreamingEwm {
    pub(crate) const fn new(alpha: f64, window_size: usize) -> Self {
        Self {
            alpha,
            current_mean: None,
            window_size,
            count: 0,
        }
    }
}

impl StreamingIndicator for StreamingEwm {
    type Input = f64;
    type Output<'a> = Option<f64>;

    fn update(&mut self, value: Self::Input) -> Self::Output<'_> {
        self.count += 1;

        match self.current_mean {
            None => {
                // First trade: initialize the mean
                self.current_mean = Some(value);
            }
            Some(prev) => {
                // Recursive update
                self.current_mean = Some((1.0 - self.alpha).mul_add(prev, self.alpha * value));
            }
        }

        if self.count >= self.window_size {
            self.current_mean
        } else {
            None
        }
    }

    fn reset(&mut self) {
        self.current_mean = None;
        self.count = 0;
    }
}

// ================================================================================================
// SMA: Simple Moving Average
// ================================================================================================

#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct StreamingSma {
    window_size: usize,
    buffer: VecDeque<f64>,
    sum: f64,
}

impl StreamingSma {
    #[must_use]
    pub fn new(window_size: SmaWindow) -> Self {
        let size = window_size.0 as usize;
        Self {
            window_size: size,
            buffer: VecDeque::with_capacity(size + 1),
            sum: 0.0,
        }
    }
}

impl StreamingIndicator for StreamingSma {
    type Input = f64;
    type Output<'a> = Option<f64>;

    #[expect(
        clippy::expect_used,
        reason = "the ring-buffer length is bounded by the window size, which always fits in u32"
    )]
    fn update(&mut self, value: Self::Input) -> Self::Output<'_> {
        self.buffer.push_back(value);
        self.sum += value;

        if self.buffer.len() > self.window_size
            && let Some(removed) = self.buffer.pop_front()
        {
            self.sum -= removed;
        }

        if self.buffer.len() >= self.window_size {
            let len_u32 =
                u32::try_from(self.buffer.len()).expect("SMA window length exceeds u32 range");
            Some(self.sum / f64::from(len_u32))
        } else {
            None
        }
    }

    fn reset(&mut self) {
        self.buffer.clear();
        self.sum = 0.0;
    }
}

// ================================================================================================
// EMA: Exponential Moving Average
// ================================================================================================

#[derive(Debug, Clone, Copy, Serialize, Deserialize)]
pub struct StreamingEma {
    inner: StreamingEwm,
}

impl StreamingEma {
    /// Creates a streaming EMA with the given window size.
    ///
    /// # Panics
    /// Panics if `window_size` cannot be represented as `u32`.
    #[must_use]
    pub fn new(window_size: EmaWindow) -> Self {
        let size = window_size.0 as usize;
        // Standard EMA Alpha = 2 / (Span + 1)
        let size_u32 = u32::from(window_size.0);
        let alpha = 2.0 / (f64::from(size_u32) + 1.0);
        Self {
            inner: StreamingEwm::new(alpha, size),
        }
    }
}

impl StreamingIndicator for StreamingEma {
    type Input = f64;
    type Output<'a> = Option<f64>;

    fn update(&mut self, value: Self::Input) -> Self::Output<'_> {
        self.inner.update(value)
    }

    fn reset(&mut self) {
        self.inner.reset();
    }
}

#[cfg(test)]
mod tests {
    #![expect(
        clippy::unwrap_used,
        reason = "tests assert against known-valid fixtures; unwrap surfaces failures as panics that fail the test"
    )]
    use super::*;

    /// Compare two floats with a small tolerance (EMA math isn't exact).
    fn approx(a: f64, b: f64) {
        assert!((a - b).abs() < 1e-9, "expected {b}, got {a}");
    }

    // ==========================================
    // SMA: Simple Moving Average
    // ==========================================

    #[test]
    fn sma_returns_none_until_window_is_full() {
        let mut sma = StreamingSma::new(SmaWindow(3));
        assert_eq!(sma.update(1.0), None); // 1 value
        assert_eq!(sma.update(2.0), None); // 2 values
        assert_eq!(sma.update(3.0), Some(2.0)); // full: (1+2+3)/3
    }

    #[test]
    fn sma_slides_the_window() {
        let mut sma = StreamingSma::new(SmaWindow(3));
        sma.update(1.0);
        sma.update(2.0);
        sma.update(3.0); // [1,2,3] -> 2.0
        assert_eq!(sma.update(4.0), Some(3.0)); // [2,3,4] -> 3.0
        assert_eq!(sma.update(5.0), Some(4.0)); // [3,4,5] -> 4.0
    }

    #[test]
    fn sma_window_of_one_returns_each_value() {
        let mut sma = StreamingSma::new(SmaWindow(1));
        assert_eq!(sma.update(7.0), Some(7.0));
        assert_eq!(sma.update(9.0), Some(9.0));
    }

    #[test]
    fn sma_reset_clears_state() {
        let mut sma = StreamingSma::new(SmaWindow(2));
        sma.update(10.0);
        sma.update(20.0); // Some(15.0)
        sma.reset();
        // Back to needing a full window again, and old values are gone.
        assert_eq!(sma.update(1.0), None);
        assert_eq!(sma.update(3.0), Some(2.0)); // (1+3)/2, not influenced by 10/20
    }

    // ==========================================
    // EMA: Exponential Moving Average
    // ==========================================

    #[test]
    fn ema_returns_none_until_window_is_reached() {
        let mut ema = StreamingEma::new(EmaWindow(3));
        assert_eq!(ema.update(1.0), None);
        assert_eq!(ema.update(2.0), None);
        assert!(ema.update(3.0).is_some()); // count >= window_size
    }

    #[test]
    fn ema_seeds_with_first_value() {
        // window 1 emits immediately; the first output is just the first input.
        let mut ema = StreamingEma::new(EmaWindow(1));
        assert_eq!(ema.update(5.0), Some(5.0));
    }

    #[test]
    fn ema_applies_alpha_recursively() {
        // alpha = 2 / (2 + 1) = 0.666...
        // v1=10 seeds mean=10 (no output yet, count 1 < 2)
        // v2=20 -> 0.6667*20 + 0.3333*10 = 16.6667 (count 2 >= 2 -> emitted)
        let mut ema = StreamingEma::new(EmaWindow(2));
        let alpha = 2.0 / 3.0;
        assert_eq!(ema.update(10.0), None);
        let expected = (1.0_f64 - alpha).mul_add(10.0, alpha * 20.0);
        approx(ema.update(20.0).unwrap(), expected);
    }

    #[test]
    fn ema_constant_input_converges_to_that_constant() {
        // Feeding the same value forever keeps the mean at that value.
        let mut ema = StreamingEma::new(EmaWindow(5));
        let mut last = None;
        for _ in 0..20 {
            last = ema.update(42.0);
        }
        approx(last.unwrap(), 42.0);
    }

    #[test]
    fn ema_reset_clears_state() {
        let mut ema = StreamingEma::new(EmaWindow(2));
        ema.update(100.0);
        ema.update(200.0); // emits something
        ema.reset();
        // After reset, the next value re-seeds from scratch.
        assert_eq!(ema.update(1.0), None); // count 1 < 2 again
        approx(
            ema.update(3.0).unwrap(),
            (1.0_f64 / 3.0).mul_add(1.0, (2.0 / 3.0) * 3.0),
        );
    }

    // ==========================================
    // EWM (shared core)
    // ==========================================

    #[test]
    fn ewm_with_alpha_one_tracks_latest_value() {
        // alpha = 1 means "no memory": output is always the newest input.
        let mut ewm = StreamingEwm::new(1.0, 1);
        assert_eq!(ewm.update(5.0), Some(5.0));
        assert_eq!(ewm.update(9.0), Some(9.0));
        assert_eq!(ewm.update(2.0), Some(2.0));
    }

    #[test]
    fn ewm_with_alpha_zero_holds_first_value() {
        // alpha = 0 means "all memory": mean stays at the seed forever.
        let mut ewm = StreamingEwm::new(0.0, 1);
        assert_eq!(ewm.update(7.0), Some(7.0));
        assert_eq!(ewm.update(100.0), Some(7.0));
    }

    #[test]
    fn ewm_respects_window_warmup() {
        let mut ewm = StreamingEwm::new(0.5, 3);
        assert_eq!(ewm.update(1.0), None);
        assert_eq!(ewm.update(2.0), None);
        assert!(ewm.update(3.0).is_some());
    }
}