Dataset and source code for "Monthly Precipitation Prediction Based on Deep Analogue and BP Ensemble"

1. Deep Analog Network (DA) 1.1zenodo\DA EBP\DA\train_whole.py: Training the DA model. Requires reading two datasets: zenodo_data\DA EBP\DA\grid\06_1951\observations5122.nc (sum of June precipitation across all grid points in the middle and lower reaches of the Yangtze River region from 1951 to 2022 (72 years), consisting of 72 values), and zenodo\DA EBP\DA\06_1951\forecasts.nc (May monthly means of 88 COI indices from 1951 to 2022 (72 years), with dimensions 88 × 1 × 72). 1.2zenodo\DA EBP\DA\predict_whole.py: Generating DL predictions. Requires reading zenodo\DA EBP\DA\min_max_value.txt and COI files in the zenodo\DA EBP\DA\coi directory, with one file per year. Each file contains 89 rows (88 factors) and 13 columns (12 months). 2. Ensemble Backpropagation Neural Network (EBP) 2.1zenodo\DA EBP\EBP\bp--50-1.txt: Using MATLAB to obtain 50 BP models.Requires two input datasets: zenodo\DA EBP\EBP\5122.txt: May data from 1951 to 2022 (72 years) consisting of 20 LFs. The file contains 72 rows and 21 columns; each row corresponds to a specific year, where column 1 is the sum of June precipitation across all grid points in the region, and columns 2 to 21 are the 20 LFs. zenodo\DA EBP\EBP\predict_5122.txt: May data from 2023 to 2025 (3 years) consisting of 20 LFs. 2.2zenodo\DA EBP\EBP\1_read.txt: Using MATLAB to generate three files: bp_2023.txt, bp_2024.txt, and bp_2025.txt. Each file contains 51 rows, corresponding to 51 ensemble members. 2.3zenodo\DA EBP\EBP\fig\: Plotting using MATLAB. 3. XAI (Explainable Artificial Intelligence) 3.1 zenodo\XAI\grid_DA\train_grid.py: Training DA for each grid point. zenodo\XAI\grid_DA\predict_grid.py: Generating LFs for each grid point across different years. 3.2 zenodo\XAI\XAI_dl.py, XAI_fal.py, XAI_gsl.py, XAI_igl.py, XAI_ig_ntl.py: Computation using explainable AI methods. zenodo\XAI\2read.py, 3read.py: Data processing. 3.3zenodo\XAI\4--fig.py: Plotting.