Given random variable X, the survival function is defined:
tfd_log_survival_function(x) = Log[ P[X > x] ] = Log[ 1 - P[X <= x] ] = Log[ 1 - cdf(x) ]
tfd_log_survival_function(distribution, value, ...)
| distribution | The distribution being used. |
|---|---|
| value | float or double Tensor. |
| ... | Additional parameters passed to Python. |
a Tensor of shape sample_shape(x) + self$batch_shape with values of type self$dtype.
Typically, different numerical approximations can be used for the log survival function, which are more accurate than 1 - cdf(x) when x >> 1.
Other distribution_methods:
tfd_cdf(),
tfd_covariance(),
tfd_cross_entropy(),
tfd_entropy(),
tfd_kl_divergence(),
tfd_log_cdf(),
tfd_log_prob(),
tfd_mean(),
tfd_mode(),
tfd_prob(),
tfd_quantile(),
tfd_sample(),
tfd_stddev(),
tfd_survival_function(),
tfd_variance()
# \donttest{ d <- tfd_normal(loc = c(1, 2), scale = c(1, 0.5)) x <- d %>% tfd_sample() d %>% tfd_log_survival_function(x)#> tf.Tensor([-0.08539618 -2.3711767 ], shape=(2,), dtype=float32)# }