https://github.com/borealisai/dcf
https://github.com/borealisai/dcf
Last synced: about 1 year ago
JSON representation
- Host: GitHub
- URL: https://github.com/borealisai/dcf
- Owner: BorealisAI
- License: other
- Created: 2024-02-14T18:52:44.000Z (over 2 years ago)
- Default Branch: main
- Last Pushed: 2024-02-14T22:55:29.000Z (over 2 years ago)
- Last Synced: 2025-02-16T03:36:11.348Z (over 1 year ago)
- Language: Python
- Size: 37.1 MB
- Stars: 1
- Watchers: 3
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# DynaConF
This is the code accompying the paper "DynaConF: Dynamic Forecasting of Non-Stationary Time Series".
## Dependencies
Install dependencies:
```bash
conda env create --file environment.yml
```
## Synthetic Data Experiments
Generate synthetic data:
```bash
run/generate_synthetic.sh
```
Run univariate baselines:
```bash
run/synthetic_baselines.sh
```
Run multivariate baselines:
```bash
run/synthetic_baselines_mv.sh
```
Run our models NAR (StatiConF) and NNAR (DynaConF):
```bash
run/synthetic_our.sh
```
Generate the result tables
```bash
run/table_synthetic.sh
```
Results are stored in `./output/synthetic/`.
## Real-World Data (Set 1) Experiments
All the real-world datasets in Set 1 are from GluonTS.
Run our models NAR (StatiConF) and NNAR (DynaConF):
```bash
run/benchmark_our_static.sh
```
and then
```bash
run/benchmark_our_dynamic.sh
```
Generate the result tables
```bash
run/table_benchmark.sh
```
Results are stored in `./output/benchmark/`.
## Real-World Data (Set 2) Experiments
All the real-world datasets in Set 2 are publically available. Information of these datasets are in `./datasets/licenses.csv`. We also include the processed datasets in `./datasets/`, which can be used by copying the unzipped folder to `~/.mxnet/gluon-ts/datasets/`.
Run our models NAR (StatiConF) and NNAR (DynaConF):
```bash
run/benchmark_new_our_static.sh
```
and then
```bash
run/benchmark_new_our_dynamic.sh
```
Run the baseslines
```bash
run/benchmark_new_baselines.sh
```
Generate the result tables
```bash
run/table_benchmark_new.sh
```
Results are stored in `./output/benchmark/`.