displacement-forecasting-toolbox-arcgispro
An interdisciplinary integration of geospatial and DL techniques underscores the potential for advanced monitoring of Earth surface dynamics.
https://github.com/lama-nm/displacement-forecasting-toolbox-arcgispro
Science Score: 57.0%
This score indicates how likely this project is to be science-related based on various indicators:
-
✓CITATION.cff file
Found CITATION.cff file -
✓codemeta.json file
Found codemeta.json file -
✓.zenodo.json file
Found .zenodo.json file -
✓DOI references
Found 2 DOI reference(s) in README -
○Academic publication links
-
○Academic email domains
-
○Institutional organization owner
-
○JOSS paper metadata
-
○Scientific vocabulary similarity
Low similarity (7.4%) to scientific vocabulary
Repository
An interdisciplinary integration of geospatial and DL techniques underscores the potential for advanced monitoring of Earth surface dynamics.
Basic Info
- Host: GitHub
- Owner: Lama-NM
- License: mit
- Default Branch: main
- Size: 7.81 KB
Statistics
- Stars: 0
- Watchers: 0
- Forks: 0
- Open Issues: 0
- Releases: 1
Metadata Files
README.md
Displacement Forecasting Toolbox for ArcGIS Pro
This Python-based toolbox enables the prediction of regular and irregular displacement time series through deep learning models, fully integrated into the ArcGIS Pro environment.
Features and Input Parameters
- Dataset File: Upload your time series CSV.
- CNR P-SBAS Checkbox: Check if your dataset originates from G-TEP (CNR).
- Output Directory: Specify where results are saved (auto-creates
output/folder). - Geographic Bounds: Input min/max latitude and longitude to define the area of interest.
- Time Steps: Define the number of future steps to predict (positive integers only).
- Optional Hyperparameters (automatically assigned if left empty):
- Number of nodes in the first LSTM layer: Defaults to 50 for 1-step prediction in both regular and irregular datasets. Defaults to 100 for multi-step prediction in regular datasets, and 8 in irregular datasets.
- Number of nodes in the second LSTM layer: Only applies to multi-step regular datasets. Defaults to 50 if unspecified.
- Learning rate:
Defaults to 0.001 for 1-step prediction in all cases.
Defaults to 0.0001 for multi-step regular datasets; remains 0.001 for irregular.
- Number of training epochs (epochs): Defaults to 15 for 1-step regular datasets and 35 for 1-step irregular. Defaults to 35 for multi-step regular datasets and 50 for multi-step irregular.
- (Optional) Plot Learning Curves: Visualize the model’s training and validation performance over epochs using RMSE as the evaluation metric.
Model Selection Logic
Depending on the dataset type and number of prediction steps:
- Regular, 1 step: LSTM model with 1 LSTM layer
- Regular, >1 steps: LSTM model with 2 LSTM layers
- Irregular, 1 step: TimeGated LSTM model
- Irregular, >1 steps: Temporal Fusion Transformer model
Output
- A shapefile containing displacement points.
- A popup showing model RMSE.
- Optional learning curves (if checkbox is enabled).
- A Tkinter GUI for interactively visualizing predictions.
Citation
Moualla, L. (2025). Displacement Forecasting Toolbox: A GIS-Integrated Deep Learning Framework. GitHub Repository. https://doi.org/10.5281/zenodo.15502468
Owner
- Login: Lama-NM
- Kind: user
- Repositories: 1
- Profile: https://github.com/Lama-NM
Citation (CITATION.cff)
cff-version: 1.2.0
message: "If you use this software, please cite it as below."
authors:
- family-names: Moualla
given-names: Lama
title: "Time Series Displacement Forecasting Toolbox"
version: "1.0.0"
date-released: 2025-05-23
url: "https://github.com/Lama-NM/displacement-forecasting-toolbox-ArcGISPRO.git"
doi: "10.5281/zenodo.15502468"
GitHub Events
Total
- Release event: 1
- Public event: 2
- Push event: 2
Last Year
- Release event: 1
- Public event: 2
- Push event: 2