Experiment Cases
Experiment Cases
layout: default title: Available num_case Values —
This page describes the different values of the num_case parameter that can be used to select the experimental case in the traffic scenario generation project.
What is num_case?
num_case is a numeric parameter that allows you to choose among various predefined cases for running experiments. Each case corresponds to a specific configuration, such as the dataset used, the model, and other relevant settings.
List of Available Studies Cases
METR-LA DATASET
Model: AE+GAN
| num_case | Case | Model Case |
|---|---|---|
| 1 | PEMS_METR | AE>GAN_linear_pretrained_METR_16 |
| 2 | PEMS_METR | AE>GAN_linear_pretrained_METR_32 |
| 3 | PEMS_METR | AE>GAN_linear_pretrained_METR_48 |
| 4 | PEMS_METR | AE>GAN_linear_pretrained_METR_64 |
PEMS_BAY DATASET
Model: AE+GAN
| num_case | Case | Model Case |
|---|---|---|
| 5 | PEMS_METR | AE>GAN_linear_pretrained_PEMS_16 |
| 6 | PEMS_METR | AE>WGAN_linear_pretrained_PEMS_16 |
| 7 | PEMS_METR | AE>GAN_linear_pretrained_PEMS_32 |
| 8 | PEMS_METR | AE>WGAN_linear_pretrained_PEMS_32 |
| 9 | PEMS_METR | AE>GAN_linear_pretrained_PEMS_48 |
| 10 | PEMS_METR | AE>GAN_linear_pretrained_PEMS_64 |
CHENGDU DATASET
Model: AE+GAN Fully Connected
| num_case | Case | Model Case |
|---|---|---|
| 11 | CHENGDU_SMALLGRAPH | AE>GAN_CHENGDU_SMALLGRAPH_16_A_linear |
| 12 | CHENGDU_SMALLGRAPH | AE>GAN_CHENGDU_SMALLGRAPH_32_A_linear |
| 13 | CHENGDU_SMALLGRAPH | AE>GAN_CHENGDU_SMALLGRAPH_48_A_linear |
| 14 | CHENGDU_SMALLGRAPH | AE>GAN_CHENGDU_SMALLGRAPH_64_A_linear |
| 15 | CHENGDU_SMALLGRAPH | AE>GAN_CHENGDU_SMALLGRAPH_96_A_linear |
| 16 | CHENGDU_SMALLGRAPH | AE>GAN_CHENGDU_SMALLGRAPH_128_A_linear |
| 17 | CHENGDU_SMALLGRAPH | AE>GAN_CHENGDU_SMALLGRAPH_192_A_linear |
| 18 | CHENGDU_SMALLGRAPH | AE>GAN_CHENGDU_SMALLGRAPH_256_A_linear |
| 19 | CHENGDU_SMALLGRAPH | AE>GAN_CHENGDU_SMALLGRAPH_512_A_linear |
Model: AE+GAN with GraphConvolutional layers
| num_case | Case | Model Case |
|---|---|---|
| gvae_0016 | CHENGDU_SMALLGRAPH | AE>GAN_CHENGDU_SMALLGRAPH_16_A_graph |
| gvae_0032 | CHENGDU_SMALLGRAPH | AE>GAN_CHENGDU_SMALLGRAPH_32_A_graph |
| gvae_0048 | CHENGDU_SMALLGRAPH | AE>GAN_CHENGDU_SMALLGRAPH_48_A_graph |
| gvae_0064 | CHENGDU_SMALLGRAPH | AE>GAN_CHENGDU_SMALLGRAPH_64_A_graph |
| gvae_0096 | CHENGDU_SMALLGRAPH | AE>GAN_CHENGDU_SMALLGRAPH_96_A_graph |
| gvae_0128 | CHENGDU_SMALLGRAPH | AE>GAN_CHENGDU_SMALLGRAPH_128_A_graph |
| gvae_0192 | CHENGDU_SMALLGRAPH | AE>GAN_CHENGDU_SMALLGRAPH_192_A_graph |
| gvae_0256 | CHENGDU_SMALLGRAPH | AE>GAN_CHENGDU_SMALLGRAPH_256_A_graph |
| gvae_0512 | CHENGDU_SMALLGRAPH | AE>GAN_CHENGDU_SMALLGRAPH_512_A_graph |
| gvae_1024 | CHENGDU_SMALLGRAPH | AE>GAN_CHENGDU_SMALLGRAPH_1024_A_graph |
Model: VAE with GraphConvolutional layers
| num_case | Case | Model Case |
|---|---|---|
| vae_0016 | CHENGDU_SMALLGRAPH | VAE_CHENGDU_SMALLGRAPH_16_A_graph_kl_fix |
| vae_0032 | CHENGDU_SMALLGRAPH | VAE_CHENGDU_SMALLGRAPH_32_A_graph_kl_fix |
| vae_0048 | CHENGDU_SMALLGRAPH | VAE_CHENGDU_SMALLGRAPH_48_A_graph_kl_fix |
| vae_0064 | CHENGDU_SMALLGRAPH | VAE_CHENGDU_SMALLGRAPH_64_A_graph_kl_fix |
| vae_0096 | CHENGDU_SMALLGRAPH | VAE_CHENGDU_SMALLGRAPH_96_A_graph_kl_fix |
| vae_0128 | CHENGDU_SMALLGRAPH | VAE_CHENGDU_SMALLGRAPH_128_A_graph_kl_fix |
| vae_0192 | CHENGDU_SMALLGRAPH | VAE_CHENGDU_SMALLGRAPH_192_A_graph_kl_fix |
| vae_0256 | CHENGDU_SMALLGRAPH | VAE_CHENGDU_SMALLGRAPH_256_A_graph_kl_fix |
| vae_0512 | CHENGDU_SMALLGRAPH | VAE_CHENGDU_SMALLGRAPH_512_A_graph_kl_fix |
| vae_1024 | CHENGDU_SMALLGRAPH | VAE_CHENGDU_SMALLGRAPH_1024_A_graph_kl_fix |
Model: B-VAE with GraphConvolutional layers B-VAE with B funciotion: linear from 0 to 1
| num_case | Case | Model Case |
|---|---|---|
| bvae_0016_lin | CHENGDU_SMALLGRAPH | VAE_CHENGDU_SMALLGRAPH_16_A_graph_kl_lin |
| bvae_0032_lin | CHENGDU_SMALLGRAPH | VAE_CHENGDU_SMALLGRAPH_32_A_graph_kl_lin |
| bvae_0048_lin | CHENGDU_SMALLGRAPH | VAE_CHENGDU_SMALLGRAPH_48_A_graph_kl_lin |
| bvae_0064_lin | CHENGDU_SMALLGRAPH | VAE_CHENGDU_SMALLGRAPH_64_A_graph_kl_lin |
| bvae_0096_lin | CHENGDU_SMALLGRAPH | VAE_CHENGDU_SMALLGRAPH_96_A_graph_kl_lin |
| bvae_0128_lin | CHENGDU_SMALLGRAPH | VAE_CHENGDU_SMALLGRAPH_128_A_graph_kl_lin |
| bvae_0192_lin | CHENGDU_SMALLGRAPH | VAE_CHENGDU_SMALLGRAPH_192_A_graph_kl_lin |
B-VAE with B funciotion: cos with period 20 epochs
| num_case | Case | Model Case |
|---|---|---|
| bvae_0016_cos | CHENGDU_SMALLGRAPH | VAE_CHENGDU_SMALLGRAPH_16_A_graph_kl_cos |
| bvae_0032_cos | CHENGDU_SMALLGRAPH | VAE_CHENGDU_SMALLGRAPH_32_A_graph_kl_cos |
| bvae_0048_cos | CHENGDU_SMALLGRAPH | VAE_CHENGDU_SMALLGRAPH_48_A_graph_kl_cos |
| bvae_0064_cos | CHENGDU_SMALLGRAPH | VAE_CHENGDU_SMALLGRAPH_64_A_graph_kl_cos |
| bvae_0096_cos | CHENGDU_SMALLGRAPH | VAE_CHENGDU_SMALLGRAPH_96_A_graph_kl_cos |
| bvae_0128_cos | CHENGDU_SMALLGRAPH | VAE_CHENGDU_SMALLGRAPH_128_A_graph_kl_cos |
| bvae_0192_cos | CHENGDU_SMALLGRAPH | VAE_CHENGDU_SMALLGRAPH_192_A_graph_kl_cos |
How to Use
To run an experiment with a specific case, use the --num_case parameter with the desired value. For example:
```bash python3 test.py –num_case 1 –experiment_name_suffix METR16_exp –main_folder experiments –repeat 5 –optimization yes –load_model no –train_models yes