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
use crate::core::Method;
use crate::core::{Error, PeriodType, ValueType, Window};
use crate::helpers::Peekable;

#[cfg(feature = "serde")]
use serde::{Deserialize, Serialize};

/// Moving [Standard Deviation](https://en.wikipedia.org/wiki/Standard_deviation) over the window of size `length` for timeseries of type [`ValueType`]
///
/// # Parameters
///
/// Has a single parameter `length`: [`PeriodType`]
///
/// `length` should be > `1`
///
/// # Input type
///
/// Input type is [`ValueType`]
///
/// # Output type
///
/// Output type is [`ValueType`]
///
/// # Examples
///
/// ```
/// use yata::prelude::*;
/// use yata::methods::StDev;
///
/// // StDev over the window with length=3
/// let mut stdev = StDev::new(3, &1.0).unwrap();
///
/// stdev.next(&1.0);
/// stdev.next(&2.0);
///
/// assert_eq!(stdev.next(&3.0), 1.0);
/// assert_eq!(stdev.next(&4.0), 1.0);
/// ```
///
/// # Performance
///
/// O(1)
///
/// [`ValueType`]: crate::core::ValueType
/// [`PeriodType`]: crate::core::PeriodType
#[derive(Debug, Clone)]
#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))]
pub struct StDev {
	mean: ValueType,
	val_sum: ValueType,
	sq_val_sum: ValueType,
	divider: ValueType,
	k: ValueType,
	window: Window<ValueType>,
}

impl Method for StDev {
	type Params = PeriodType;
	type Input = ValueType;
	type Output = Self::Input;

	fn new(length: Self::Params, &value: &Self::Input) -> Result<Self, Error> {
		match length {
			0 | 1 => Err(Error::WrongMethodParameters),
			length => {
				let k = ((length - 1) as ValueType).recip();

				let float_length = length as ValueType;
				let mean = -value;
				let divider = -float_length.recip();

				Ok(Self {
					mean,
					val_sum: value * float_length,
					sq_val_sum: value * value * float_length,
					divider,
					k,
					window: Window::new(length, value),
				})
			}
		}
	}

	#[inline]
	fn next(&mut self, &value: &Self::Input) -> Self::Output {
		let prev_value = self.window.push(value);
		let diff = value - prev_value;

		// same as `value * value - prev_value * prev_value`
		self.sq_val_sum += diff * (value + prev_value);

		self.val_sum += diff;
		self.mean += diff * self.divider;

		self.peek()
	}
}

impl Peekable<<Self as Method>::Output> for StDev {
	fn peek(&self) -> <Self as Method>::Output {
		// self.sq_val_sum - self.val_sum * self.mean;
		let sum = self.val_sum.mul_add(self.mean, self.sq_val_sum);

		(sum * self.k)
			.abs() // sometimes float values may produce negative values, when sum is really near to zero value
			.sqrt()
	}
}

#[cfg(test)]
#[allow(clippy::suboptimal_flops)]
mod tests {
	use super::{Method, StDev as TestingMethod};
	use crate::core::ValueType;
	use crate::helpers::{assert_eq_float, RandomCandles};
	use crate::methods::tests::test_const_float;

	#[test]
	fn test_st_dev_const() {
		for i in 2..255 {
			let input = (i as ValueType + 56.0) / 16.3251;
			let mut method = TestingMethod::new(i, &input).unwrap();

			test_const_float(&mut method, &input, 0.0);
		}
	}

	#[test]
	fn test_st_dev() {
		let candles = RandomCandles::default();

		let src: Vec<ValueType> = candles
			.take(300)
			.enumerate()
			.map(|(i, x)| x.close * if i % 2 == 0 { 1.0 } else { -1.0 })
			.collect();

		(2..255).for_each(|ma_length| {
			let mut ma = TestingMethod::new(ma_length, &src[0]).unwrap();
			let ma_length = ma_length as usize;

			src.iter().enumerate().for_each(|(i, x)| {
				let mut avg = 0.;
				for j in 0..ma_length {
					avg += src[i.saturating_sub(j)] / ma_length as ValueType;
				}

				let mut diff_sq_sum = 0.;
				for j in 0..ma_length {
					diff_sq_sum +=
						(src[i.saturating_sub(j)] - avg).powi(2) / (ma_length - 1) as ValueType;
				}

				let value = ma.next(x);
				let value2 = diff_sq_sum.sqrt();

				println!("{value2} <=> {value} at {i} with length {ma_length}");
				assert_eq_float(value2, value);
			});
		});
	}
}