MindElec
mindelec.architecture
API Name |
Description |
Supported Platforms |
The dense layer sequential. |
|
|
Scale the input value to specified region. |
|
|
The LinearBlock. |
|
|
The MTL strategy weighted multi-task losses automatically. |
|
|
The multi-scale network. |
|
|
The ResBlock of dense layer. |
|
|
Gets the activation function. |
|
mindelec.common
API Name |
Description |
Supported Platforms |
Calculates l2 metric. |
|
|
Warmup-decay learning rate. |
|
|
generate learning rate array |
|
mindelec.data
API Name |
Description |
Supported Platforms |
Bounding box for sampling space, only supports cube-shape sampling space, at present supports STATIC(0) and DYNAMIC(1). |
|
|
Sampling data of boundary condition. |
|
|
Sampling data of initial condition. |
|
|
Combine datasets together. |
|
|
Sampling data of equation domain. |
|
|
Set arguments of ExistedDataset. |
|
|
Creates a dataset with given data path. |
|
|
Material solution config for PointCloud-Tensor generation, which influence the material solving stage. |
|
|
Read the stp files to generate PointCloud data, for downstream physical-equation AI simulation. |
|
|
Sampling space config for PointCloud-Tensor generation. |
|
|
Point sampling method, at present support UPPERBOUND(0) and DIMENSIONS(1). |
|
|
Standard physical quantities fields that Maxwell equations concern about, material solving stage will deal with these standard physical fields. |
|
mindelec.geometry
API Name |
Description |
Supported Platforms |
Convert from dict to SamplingConfig. |
|
|
CSG class for difference of geometry. |
|
|
CSG class for intersection of geometries. |
|
|
CSG class for union of geometries. |
|
|
CSG class for xor of geometries. |
|
|
Definition of Cuboid object. |
|
|
Definition of Disk object. |
|
|
Definition of Geometry object. |
|
|
Definition of geometry with time. |
|
|
Definition of HyperCube object. |
|
|
Definition of Interval object. |
|
|
Definition of partial sampling configuration. |
|
|
Definition of Rectangle object. |
|
|
Definition of global sampling configuration. |
|
|
Definition of Time Domain. |
|
mindelec.loss
API Name |
Description |
Supported Platforms |
Definition of the loss for all sub-dataset. |
|
|
Encapsulation class of network with loss of eval. |
|
|
Encapsulation class of network with loss. |
|
|
Gets the loss function. |
|
mindelec.operators
API Name |
Description |
Supported Platforms |
Computes and returns the gradients of the specified column of outputs with respect to the specified column of inputs. |
|
|
Computes and returns the second order gradients of the specified column of outputs with respect to the specified column of inputs. |
|
mindelec.solver
API Name |
Description |
Supported Platforms |
Evaluate the model during training. |
|
|
Monitor the loss in training. |
|
|
Base class of user-defined problems. |
|
|
High-Level API for training or inference. |
|
mindelec.vision
API Name |
Description |
Supported Platforms |
LossMonitor for eval. |
|
|
Loss monitor for train. |
|
|
Create video from existing images. |
|
|
Draw electric and magnetic field values of every timestep for 2D slices, and save them in path_image_save |
|
|
Draw s11-frequency curve and save it in path_image_save. |
|
|
Draw 1d scatter image |
|
|
Draw 2d scatter image |
|
|
Generates 3D vtk file for visualizaiton. |
|