gluonts.nursery.auto_ode.auto_ode module#

class gluonts.nursery.auto_ode.auto_ode.LowRankVectorEmbedding(neural_lv, **kwargs)[source]#

Bases: mxnet.gluon.block.HybridBlock

forward()[source]#: Defines the forward computation. Arguments can be either NDArray or Symbol.

class gluonts.nursery.auto_ode.auto_ode.NeuralLV(num_time_series, num_time_steps, low_rank_param, is_full_matrix, p0, r, k, A, is_sym)[source]#

Bases: object

run(num_epochs=1000, model=None)[source]#

solve_discrete_lv(mat1, mat2=None, is_full_matrix=True)[source]#

gluonts.nursery.auto_ode.auto_ode.compute_low_rank_product(B, C)[source]#

gluonts.nursery.auto_ode.auto_ode.compute_mat_vec_prod(B, C, p)[source]#

gluonts.nursery.auto_ode.auto_ode.generate_data(num_ts, ctx=gpu(0), dtype='float64', seed=100)[source]#

gluonts.nursery.auto_ode.auto_ode.lv_plot_ts(p, p_approx, max_num_plots=10, num_rows=2, fig_size_width=10)[source]#

gluonts.nursery.auto_ode.auto_ode.train(p, trainer, model, num_epochs=1000)[source]#