Spaces:
Sleeping
Sleeping
File size: 8,506 Bytes
be7c937 | 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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 | #include "wayy_db/ops/window.hpp"
#include <deque>
#include <cmath>
#include <numeric>
namespace wayy_db::ops {
// Moving average
std::vector<double> mavg(const ColumnView<double>& col, size_t window) {
if (col.empty() || window == 0) return {};
std::vector<double> result(col.size());
double sum = 0.0;
for (size_t i = 0; i < col.size(); ++i) {
sum += col[i];
if (i >= window) {
sum -= col[i - window];
result[i] = sum / static_cast<double>(window);
} else {
result[i] = sum / static_cast<double>(i + 1);
}
}
return result;
}
std::vector<double> mavg(const ColumnView<int64_t>& col, size_t window) {
if (col.empty() || window == 0) return {};
std::vector<double> result(col.size());
int64_t sum = 0;
for (size_t i = 0; i < col.size(); ++i) {
sum += col[i];
if (i >= window) {
sum -= col[i - window];
result[i] = static_cast<double>(sum) / static_cast<double>(window);
} else {
result[i] = static_cast<double>(sum) / static_cast<double>(i + 1);
}
}
return result;
}
// Moving sum
std::vector<double> msum(const ColumnView<double>& col, size_t window) {
if (col.empty() || window == 0) return {};
std::vector<double> result(col.size());
double sum = 0.0;
for (size_t i = 0; i < col.size(); ++i) {
sum += col[i];
if (i >= window) {
sum -= col[i - window];
}
result[i] = sum;
}
return result;
}
std::vector<int64_t> msum(const ColumnView<int64_t>& col, size_t window) {
if (col.empty() || window == 0) return {};
std::vector<int64_t> result(col.size());
int64_t sum = 0;
for (size_t i = 0; i < col.size(); ++i) {
sum += col[i];
if (i >= window) {
sum -= col[i - window];
}
result[i] = sum;
}
return result;
}
// Moving standard deviation (Welford's online algorithm)
std::vector<double> mstd(const ColumnView<double>& col, size_t window) {
if (col.empty() || window == 0) return {};
std::vector<double> result(col.size());
for (size_t i = 0; i < col.size(); ++i) {
size_t start = (i >= window) ? i - window + 1 : 0;
size_t count = i - start + 1;
double mean = 0.0;
double m2 = 0.0;
size_t n = 0;
for (size_t j = start; j <= i; ++j) {
++n;
double delta = col[j] - mean;
mean += delta / static_cast<double>(n);
double delta2 = col[j] - mean;
m2 += delta * delta2;
}
result[i] = (n > 1) ? std::sqrt(m2 / static_cast<double>(n)) : 0.0;
}
return result;
}
std::vector<double> mstd(const ColumnView<int64_t>& col, size_t window) {
if (col.empty() || window == 0) return {};
std::vector<double> result(col.size());
for (size_t i = 0; i < col.size(); ++i) {
size_t start = (i >= window) ? i - window + 1 : 0;
double mean = 0.0;
double m2 = 0.0;
size_t n = 0;
for (size_t j = start; j <= i; ++j) {
++n;
double val = static_cast<double>(col[j]);
double delta = val - mean;
mean += delta / static_cast<double>(n);
double delta2 = val - mean;
m2 += delta * delta2;
}
result[i] = (n > 1) ? std::sqrt(m2 / static_cast<double>(n)) : 0.0;
}
return result;
}
// Moving min/max using monotonic deque for O(n) complexity
template<typename T, typename Compare>
std::vector<T> monotonic_window(const ColumnView<T>& col, size_t window, Compare cmp) {
if (col.empty() || window == 0) return {};
std::vector<T> result(col.size());
std::deque<size_t> dq; // Indices
for (size_t i = 0; i < col.size(); ++i) {
// Remove elements outside window
while (!dq.empty() && dq.front() + window <= i) {
dq.pop_front();
}
// Remove elements that won't be min/max
while (!dq.empty() && cmp(col[i], col[dq.back()])) {
dq.pop_back();
}
dq.push_back(i);
result[i] = col[dq.front()];
}
return result;
}
std::vector<double> mmin(const ColumnView<double>& col, size_t window) {
return monotonic_window(col, window, std::less<double>{});
}
std::vector<int64_t> mmin(const ColumnView<int64_t>& col, size_t window) {
return monotonic_window(col, window, std::less<int64_t>{});
}
std::vector<double> mmax(const ColumnView<double>& col, size_t window) {
return monotonic_window(col, window, std::greater<double>{});
}
std::vector<int64_t> mmax(const ColumnView<int64_t>& col, size_t window) {
return monotonic_window(col, window, std::greater<int64_t>{});
}
// Exponential moving average
std::vector<double> ema(const ColumnView<double>& col, double alpha) {
if (col.empty()) return {};
if (alpha <= 0.0 || alpha > 1.0) {
throw std::invalid_argument("EMA alpha must be in (0, 1]");
}
std::vector<double> result(col.size());
result[0] = col[0];
for (size_t i = 1; i < col.size(); ++i) {
result[i] = alpha * col[i] + (1.0 - alpha) * result[i - 1];
}
return result;
}
std::vector<double> ema(const ColumnView<int64_t>& col, double alpha) {
if (col.empty()) return {};
if (alpha <= 0.0 || alpha > 1.0) {
throw std::invalid_argument("EMA alpha must be in (0, 1]");
}
std::vector<double> result(col.size());
result[0] = static_cast<double>(col[0]);
for (size_t i = 1; i < col.size(); ++i) {
result[i] = alpha * static_cast<double>(col[i]) + (1.0 - alpha) * result[i - 1];
}
return result;
}
std::vector<double> ema_span(const ColumnView<double>& col, size_t span) {
double alpha = 2.0 / (static_cast<double>(span) + 1.0);
return ema(col, alpha);
}
// Diff
std::vector<double> diff(const ColumnView<double>& col, size_t periods) {
if (col.empty() || periods >= col.size()) return std::vector<double>(col.size(), 0.0);
std::vector<double> result(col.size());
for (size_t i = 0; i < periods; ++i) {
result[i] = std::numeric_limits<double>::quiet_NaN();
}
for (size_t i = periods; i < col.size(); ++i) {
result[i] = col[i] - col[i - periods];
}
return result;
}
std::vector<int64_t> diff(const ColumnView<int64_t>& col, size_t periods) {
if (col.empty() || periods >= col.size()) return std::vector<int64_t>(col.size(), 0);
std::vector<int64_t> result(col.size(), 0);
for (size_t i = periods; i < col.size(); ++i) {
result[i] = col[i] - col[i - periods];
}
return result;
}
// Percent change
std::vector<double> pct_change(const ColumnView<double>& col, size_t periods) {
if (col.empty() || periods >= col.size()) {
return std::vector<double>(col.size(), std::numeric_limits<double>::quiet_NaN());
}
std::vector<double> result(col.size());
for (size_t i = 0; i < periods; ++i) {
result[i] = std::numeric_limits<double>::quiet_NaN();
}
for (size_t i = periods; i < col.size(); ++i) {
if (col[i - periods] != 0.0) {
result[i] = (col[i] - col[i - periods]) / col[i - periods];
} else {
result[i] = std::numeric_limits<double>::quiet_NaN();
}
}
return result;
}
// Shift
std::vector<double> shift(const ColumnView<double>& col, int64_t n) {
if (col.empty()) return {};
std::vector<double> result(col.size(), std::numeric_limits<double>::quiet_NaN());
if (n >= 0) {
size_t offset = static_cast<size_t>(n);
for (size_t i = offset; i < col.size(); ++i) {
result[i] = col[i - offset];
}
} else {
size_t offset = static_cast<size_t>(-n);
for (size_t i = 0; i + offset < col.size(); ++i) {
result[i] = col[i + offset];
}
}
return result;
}
std::vector<int64_t> shift(const ColumnView<int64_t>& col, int64_t n) {
if (col.empty()) return {};
std::vector<int64_t> result(col.size(), 0);
if (n >= 0) {
size_t offset = static_cast<size_t>(n);
for (size_t i = offset; i < col.size(); ++i) {
result[i] = col[i - offset];
}
} else {
size_t offset = static_cast<size_t>(-n);
for (size_t i = 0; i + offset < col.size(); ++i) {
result[i] = col[i + offset];
}
}
return result;
}
} // namespace wayy_db::ops
|