| /third_party/mindspore/tests/ut/python/nn/probability/distribution/ |
| D | test_exponential.py | 118 def construct(self, rate_b, rate_a): argument
119 kl1 = self.e1.kl_loss('Exponential', rate_b)
120 kl2 = self.e2.kl_loss('Exponential', rate_b, rate_a)
128 rate_b = Tensor([0.3], dtype=dtype.float32)
130 ans = net(rate_b, rate_a)
142 def construct(self, rate_b, rate_a): argument
143 h1 = self.e1.cross_entropy('Exponential', rate_b)
144 h2 = self.e2.cross_entropy('Exponential', rate_b, rate_a)
152 rate_b = Tensor([0.3], dtype=dtype.float32)
154 ans = net(rate_b, rate_a)
|
| D | test_gamma.py | 131 def construct(self, concentration_b, rate_b, concentration_a, rate_a): argument
132 kl1 = self.g1.kl_loss('Gamma', concentration_b, rate_b)
133 kl2 = self.g2.kl_loss('Gamma', concentration_b, rate_b, concentration_a, rate_a)
142 rate_b = Tensor(np.array([1.0]).astype(np.float32), dtype=dtype.float32)
145 ans = net(concentration_b, rate_b, concentration_a, rate_a)
157 def construct(self, concentration_b, rate_b, concentration_a, rate_a): argument
158 h1 = self.g1.cross_entropy('Gamma', concentration_b, rate_b)
159 h2 = self.g2.cross_entropy('Gamma', concentration_b, rate_b, concentration_a, rate_a)
168 rate_b = Tensor(np.array([1.0]).astype(np.float32), dtype=dtype.float32)
171 ans = net(concentration_b, rate_b, concentration_a, rate_a)
|
| /third_party/mindspore/mindspore/nn/probability/distribution/ |
| D | exponential.py | 215 def _cross_entropy(self, dist, rate_b, rate=None): argument
225 return self._entropy(rate) + self._kl_loss(dist, rate_b, rate)
295 def _kl_loss(self, dist, rate_b, rate=None): argument
305 rate_b = self._check_value(rate_b, 'rate_b')
306 rate_b = self.cast(rate_b, self.parameter_type)
308 return self.log(rate_a) - self.log(rate_b) + rate_b / rate_a - 1.0
|
| D | gamma.py | 262 def _cross_entropy(self, dist, concentration_b, rate_b, concentration_a=None, rate_a=None): argument
275 self._kl_loss(dist, concentration_b, rate_b, concentration_a, rate_a)
314 def _kl_loss(self, dist, concentration_b, rate_b, concentration_a=None, rate_a=None): argument
332 rate_b = self._check_value(rate_b, 'rate_b')
334 rate_b = self.cast(rate_b, self.parameter_type)
338 + concentration_b * self.log(rate_a) - concentration_b * self.log(rate_b) \
339 + concentration_a * (rate_b / rate_a - 1.)
|
| /third_party/mindspore/tests/st/probability/distribution/ |
| D | test_gamma.py | 92 rate_b = np.array([1.0]).astype(np.float32)
96 + concentration_b * np.log(rate_a) - concentration_b * np.log(rate_b) \
97 + concentration_a * (rate_b / rate_a - 1.)
101 rate = Tensor(rate_b, dtype=dtype.float32)
316 rate_b = np.array([1.0]).astype(np.float32)
317 ans = prob(Tensor(concentration_b), Tensor(rate_b))
321 + concentration_b * np.log(rate_a) - concentration_b * np.log(rate_b) \
322 + concentration_a * (rate_b / rate_a - 1.)
|
| D | test_exponential.py | 88 rate_b = np.array([0.5, 2.0]).astype(np.float32)
89 expect_kl_loss = np.log(rate_a) - np.log(rate_b) + rate_b / rate_a - 1.0
91 output = kl(Tensor(rate_b, dtype=dtype.float32))
|
| /third_party/pulseaudio/src/pulsecore/ |
| D | resampler.c | 140 const uint32_t rate_b) { in fix_method() argument
143 pa_assert(pa_sample_rate_valid(rate_b)); in fix_method()
147 if (!(flags & PA_RESAMPLER_VARIABLE_RATE) && rate_a == rate_b) { in fix_method()
159 if (rate_a != rate_b) { in fix_method()
178 if (rate_a < rate_b) { in fix_method()
|