def cdf(self, x: Tensor):
# Note: computes CDF on each dimension of the target independently.
ix = [
np.searchsorted(arr, val)
for arr, val in zip(self.sorted_samples.T.asnumpy(), x.asnumpy())
]
CDF_sorted = self.F.linspace(
start=1 / len(self.samples),
stop=1,
endpoint=True,
num=len(self.samples),
)
return CDF_sorted.take(indices=mx.nd.array(ix), axis=0)
def cdf(self, x: Tensor):
# Note: computes CDF on each dimension of the target independently.
self.CDF_sorted = self.F.linspace(
start=1 / self.samples.shape[0],
stop=1,
endpoint=True,
num=self.samples.shape[0],
)
# Replace this when mxnet is updated to a version where `searchsorted` is available!
import torch
sorted_samples = torch.tensor(self.sorted_samples.asnumpy()).transpose(
0, -1)
x = torch.tensor(x.asnumpy()).unsqueeze(dim=-1)
ix = torch.searchsorted(sorted_samples, x).squeeze().numpy()
return self.CDF_sorted.take(indices=mx.nd.array(ix), axis=0)
def hybrid_forward(
self,
F,
data: Tensor,
observed_indicator: Tensor,
scale: Optional[Tensor],
rep_params: List[Tensor],
**kwargs,
) -> Tuple[Tensor, Tensor, List[Tensor]]:
data_np = data.asnumpy()
observed_indicator_np = observed_indicator.astype("int32").asnumpy()
if scale is None:
# Even though local binning implicitly scales the data, we still return the scale as an input to the model.
scale = F.expand_dims(
F.sum(data * observed_indicator, axis=-1)
/ F.sum(observed_indicator, axis=-1),
-1,
)
bin_centers_hyb = np.ones((len(data), self.num_bins)) * (-1)
bin_edges_hyb = np.ones((len(data), self.num_bins + 1)) * (-1)
# Every time series needs to be binned individually
for i in range(len(data_np)):
# Identify observed data points.
data_loc = data_np[i]
observed_indicator_loc = observed_indicator_np[i]
data_obs_loc = data_loc[observed_indicator_loc == 1]
if data_obs_loc.size > 0:
# Calculate time series specific bin centers and edges.
if self.is_quantile:
bin_centers_loc = np.quantile(
data_obs_loc, np.linspace(0, 1, self.num_bins)
)
else:
bin_centers_loc = np.linspace(
np.min(data_obs_loc),
np.max(data_obs_loc),
self.num_bins,
)
bin_centers_hyb[i] = ensure_binning_monotonicity(
bin_centers_loc
)
bin_edges_hyb[i] = bin_edges_from_bin_centers(
bin_centers_hyb[i]
)
# Bin the time series.
data_obs_loc_binned = np.digitize(
data_obs_loc, bins=bin_edges_hyb[i], right=False
)
else:
data_obs_loc_binned = []
# Write the binned time series back into the data array.
data_loc[observed_indicator_loc == 1] = data_obs_loc_binned
data_np[i] = data_loc
else:
bin_centers_hyb = rep_params[0].asnumpy()
bin_edges_hyb = rep_params[1].asnumpy()
bin_edges_hyb = np.repeat(
bin_edges_hyb,
len(data_np) // len(bin_edges_hyb),
axis=0,
)
bin_centers_hyb = np.repeat(
bin_centers_hyb,
len(data_np) // len(bin_centers_hyb),
axis=0,
)
for i in range(len(data_np)):
data_loc = data_np[i]
observed_indicator_loc = observed_indicator_np[i]
data_obs_loc = data_loc[observed_indicator_loc == 1]
# Bin the time series based on previously computed bin edges.
data_obs_loc_binned = np.digitize(
data_obs_loc, bins=bin_edges_hyb[i], right=False
)
data_loc[observed_indicator_loc == 1] = data_obs_loc_binned
data_np[i] = data_loc
bin_centers_hyb = F.array(bin_centers_hyb)
bin_edges_hyb = F.array(bin_edges_hyb)
data = mx.nd.array(data_np)
return data, scale, [bin_centers_hyb, bin_edges_hyb]
def plot_samples(s: Tensor, bins: int = 100) -> None:
from matplotlib import pyplot as plt
s = s.asnumpy()
plt.hist(s, bins=bins)
plt.show()
