# ezrah442/climate-feedback-prediction architecture

> Architecture diagram of the GitHub repository ezrah442/climate-feedback-prediction, generated by GitDiagram from its file tree, README and a sample of source files (last updated 2026-10-02). It is an AI-made overview: check details against the source.

- Interactive diagram: https://gitdiagram.com/ezrah442/climate-feedback-prediction
- Repository: https://github.com/ezrah442/climate-feedback-prediction

## Overview

This repository is a climate-science machine-learning workflow for estimating radiative responses from atmospheric inputs and evaluating the resulting models. The main path is configuration and climate-data preparation, preprocessing and dataset assembly, model training, then inference and scientific analysis of predictions, kernels, and closure residuals. Separate evaluation scripts cover ensembles, autoencoders, clouds, legacy models, and comparison tables. The README is empty; relationships below rely on the supplied source excerpts and SOURCE INDEX, so unsampled connections are omitted or marked unverified.

Main entry points: [`config_utils.py`](https://github.com/ezrah442/climate-feedback-prediction/blob/main/config_utils.py) (Configuration), [`preprocessing.py`](https://github.com/ezrah442/climate-feedback-prediction/blob/main/preprocessing.py) (Preprocessing), [`eval.py`](https://github.com/ezrah442/climate-feedback-prediction/blob/main/eval.py) (Model evaluation), [`dataloader.py`](https://github.com/ezrah442/climate-feedback-prediction/blob/main/dataloader.py) (Climate dataset)

## Diagram (Mermaid)

```mermaid
flowchart TD

subgraph group_prep["Data preparation"]
  node_data_requests["Data requests<br/>[data_requests.py]"]
  node_preprocess["Preprocessing<br/>[preprocessing.py]"]
  node_ecod["ECOD calculation"]
  node_prepare_cli["Preprocess runner<br/>[preprocess.py]"]
  node_loader["Climate dataset<br/>[dataloader.py]"]
end

subgraph group_learning["Model learning"]
  node_model["Neural model<br/>[model.py]"]
  node_train["Training runner<br/>[train.py]"]
end

subgraph group_analysis["Scientific evaluation"]
  node_eval["Model evaluation<br/>[eval.py]"]
  node_eval_ensemble["Ensemble evaluation<br/>[eval_ensemble.py]"]
  node_closure["Closure analysis<br/>[eval_closure.py]"]
  node_closure_ensemble["Ensemble closure"]
  node_comparisons["Model comparisons"]
end

subgraph group_legacy["Specialized analyses"]
  node_autoencoder["Autoencoder evaluation"]
  node_autoencoder_interactive["Interactive autoencoder"]
  node_cloud_eval["Cloud evaluation<br/>[eval_cloud.py]"]
  node_old_eval["Legacy model evaluation<br/>[eval_old_model.py]"]
  node_old_closure["Legacy closure"]
  node_shakirova["Shakirova analysis<br/>[eval_shakirova.py]"]
end

subgraph group_support["Shared runtime"]
  node_config["Configuration<br/>[config_utils.py]"]
  node_utils["Climate utilities<br/>[utils.py]"]
  node_cartopy["Map utilities<br/>[utils_cartopy.py]"]
end

node_researcher(("Researcher"))
node_climate_data[("Climate data")]

node_researcher -.->|"runs preparation"| node_prepare_cli
node_researcher -.->|"trains model"| node_train
node_researcher -.->|"runs evaluation"| node_eval
node_climate_data -.->|"loads climate inputs"| node_loader
node_prepare_cli -->|"uses preprocessing"| node_preprocess
node_prepare_cli -->|"loads settings"| node_config
node_preprocess -->|"builds from settings"| node_config
node_preprocess -->|"calculates ECOD"| node_ecod
node_loader -->|"loads settings"| node_config
node_loader -->|"transforms inputs"| node_preprocess
node_loader -->|"uses data helpers"| node_utils
node_train -.->|"loads settings"| node_config
node_train -.->|"loads batches"| node_loader
node_train -.->|"trains model"| node_model
node_utils -->|"uses variable config"| node_config
node_utils -->|"uses transforms"| node_preprocess
node_utils -->|"loads model"| node_model
node_eval -->|"loads settings"| node_config
node_eval -->|"transforms inputs"| node_preprocess
node_eval -->|"runs predictions"| node_model
node_eval -->|"plots maps"| node_cartopy
node_eval_ensemble -->|"loads settings"| node_config
node_eval_ensemble -->|"transforms inputs"| node_preprocess
node_eval_ensemble -->|"runs ensemble"| node_model
node_closure -->|"loads settings"| node_config
node_closure -->|"plots maps"| node_cartopy
node_closure_ensemble -->|"loads settings"| node_config
node_comparisons -->|"reads settings"| node_config
node_comparisons -->|"uses evaluation results"| node_eval
node_comparisons -->|"uses preprocessing results"| node_prepare_cli
node_autoencoder -->|"loads settings"| node_config
node_autoencoder_interactive -->|"uses evaluation"| node_autoencoder
node_cloud_eval -->|"loads settings"| node_config
node_old_closure -->|"uses legacy evaluation"| node_old_eval
node_old_closure -->|"uses closure analysis"| node_closure
node_old_eval -->|"uses climate helpers"| node_utils
node_shakirova -->|"loads settings"| node_config
node_shakirova -->|"plots maps"| node_cartopy

click node_config "https://github.com/ezrah442/climate-feedback-prediction/blob/main/config_utils.py"
click node_data_requests "https://github.com/ezrah442/climate-feedback-prediction/blob/main/data_requests.py"
click node_preprocess "https://github.com/ezrah442/climate-feedback-prediction/blob/main/preprocessing.py"
click node_ecod "https://github.com/ezrah442/climate-feedback-prediction/blob/main/ecod_calculation.py"
click node_prepare_cli "https://github.com/ezrah442/climate-feedback-prediction/blob/main/preprocess.py"
click node_loader "https://github.com/ezrah442/climate-feedback-prediction/blob/main/dataloader.py"
click node_model "https://github.com/ezrah442/climate-feedback-prediction/blob/main/model.py"
click node_train "https://github.com/ezrah442/climate-feedback-prediction/blob/main/train.py"
click node_eval "https://github.com/ezrah442/climate-feedback-prediction/blob/main/eval.py"
click node_eval_ensemble "https://github.com/ezrah442/climate-feedback-prediction/blob/main/eval_ensemble.py"
click node_closure "https://github.com/ezrah442/climate-feedback-prediction/blob/main/eval_closure.py"
click node_closure_ensemble "https://github.com/ezrah442/climate-feedback-prediction/blob/main/eval_closure_ensemble.py"
click node_comparisons "https://github.com/ezrah442/climate-feedback-prediction/blob/main/make_comparison_tables.py"
click node_autoencoder "https://github.com/ezrah442/climate-feedback-prediction/blob/main/eval_autoencoder.py"
click node_autoencoder_interactive "https://github.com/ezrah442/climate-feedback-prediction/blob/main/eval_autoencoder_interactive.py"
click node_cloud_eval "https://github.com/ezrah442/climate-feedback-prediction/blob/main/eval_cloud.py"
click node_old_eval "https://github.com/ezrah442/climate-feedback-prediction/blob/main/eval_old_model.py"
click node_old_closure "https://github.com/ezrah442/climate-feedback-prediction/blob/main/eval_closure_old_model.py"
click node_shakirova "https://github.com/ezrah442/climate-feedback-prediction/blob/main/eval_shakirova.py"
click node_utils "https://github.com/ezrah442/climate-feedback-prediction/blob/main/utils.py"
click node_cartopy "https://github.com/ezrah442/climate-feedback-prediction/blob/main/utils_cartopy.py"

classDef toneNeutral fill:#f8fafc,stroke:#334155,stroke-width:1.5px,color:#0f172a
classDef toneBlue fill:#dbeafe,stroke:#2563eb,stroke-width:1.5px,color:#172554
classDef toneAmber fill:#fef3c7,stroke:#d97706,stroke-width:1.5px,color:#78350f
classDef toneMint fill:#dcfce7,stroke:#16a34a,stroke-width:1.5px,color:#14532d
classDef toneRose fill:#ffe4e6,stroke:#e11d48,stroke-width:1.5px,color:#881337
classDef toneIndigo fill:#e0e7ff,stroke:#4f46e5,stroke-width:1.5px,color:#312e81
classDef toneTeal fill:#ccfbf1,stroke:#0f766e,stroke-width:1.5px,color:#134e4a
class node_data_requests,node_preprocess,node_ecod,node_prepare_cli,node_loader toneBlue
class node_model,node_train,node_climate_data toneAmber
class node_eval,node_eval_ensemble,node_closure,node_closure_ensemble,node_comparisons toneMint
class node_autoencoder,node_autoencoder_interactive,node_cloud_eval,node_old_eval,node_old_closure,node_shakirova toneRose
class node_config,node_utils,node_cartopy,node_researcher toneIndigo
```

## Components

### Data preparation

- [Data requests](https://github.com/ezrah442/climate-feedback-prediction/blob/main/data_requests.py): `data_requests.py`
- [Preprocessing](https://github.com/ezrah442/climate-feedback-prediction/blob/main/preprocessing.py): `preprocessing.py`
- [ECOD calculation](https://github.com/ezrah442/climate-feedback-prediction/blob/main/ecod_calculation.py): `ecod_calculation.py`
- [Preprocess runner](https://github.com/ezrah442/climate-feedback-prediction/blob/main/preprocess.py): `preprocess.py`
- [Climate dataset](https://github.com/ezrah442/climate-feedback-prediction/blob/main/dataloader.py): `dataloader.py`

### Model learning

- [Neural model](https://github.com/ezrah442/climate-feedback-prediction/blob/main/model.py): `model.py`
- [Training runner](https://github.com/ezrah442/climate-feedback-prediction/blob/main/train.py): `train.py`

### Scientific evaluation

- [Model evaluation](https://github.com/ezrah442/climate-feedback-prediction/blob/main/eval.py): `eval.py`
- [Ensemble evaluation](https://github.com/ezrah442/climate-feedback-prediction/blob/main/eval_ensemble.py): `eval_ensemble.py`
- [Closure analysis](https://github.com/ezrah442/climate-feedback-prediction/blob/main/eval_closure.py): `eval_closure.py`
- [Ensemble closure](https://github.com/ezrah442/climate-feedback-prediction/blob/main/eval_closure_ensemble.py): `eval_closure_ensemble.py`
- [Model comparisons](https://github.com/ezrah442/climate-feedback-prediction/blob/main/make_comparison_tables.py): `make_comparison_tables.py`

### Specialized analyses

- [Autoencoder evaluation](https://github.com/ezrah442/climate-feedback-prediction/blob/main/eval_autoencoder.py): `eval_autoencoder.py`
- [Interactive autoencoder](https://github.com/ezrah442/climate-feedback-prediction/blob/main/eval_autoencoder_interactive.py): `eval_autoencoder_interactive.py`
- [Cloud evaluation](https://github.com/ezrah442/climate-feedback-prediction/blob/main/eval_cloud.py): `eval_cloud.py`
- [Legacy model evaluation](https://github.com/ezrah442/climate-feedback-prediction/blob/main/eval_old_model.py): `eval_old_model.py`
- [Legacy closure](https://github.com/ezrah442/climate-feedback-prediction/blob/main/eval_closure_old_model.py): `eval_closure_old_model.py`
- [Shakirova analysis](https://github.com/ezrah442/climate-feedback-prediction/blob/main/eval_shakirova.py): `eval_shakirova.py`

### Shared runtime

- [Configuration](https://github.com/ezrah442/climate-feedback-prediction/blob/main/config_utils.py): `config_utils.py`
- [Climate utilities](https://github.com/ezrah442/climate-feedback-prediction/blob/main/utils.py): `utils.py`
- [Map utilities](https://github.com/ezrah442/climate-feedback-prediction/blob/main/utils_cartopy.py): `utils_cartopy.py`

### Other components

- **Researcher**
- **Climate data**

## Connections

- Researcher → Preprocess runner: runs preparation
- Researcher → Training runner: trains model
- Researcher → Model evaluation: runs evaluation
- Climate data → Climate dataset: loads climate inputs
- Preprocess runner → Preprocessing: uses preprocessing (evidence: `preprocess.py`)
- Preprocess runner → Configuration: loads settings (evidence: `preprocess.py`)
- Preprocessing → Configuration: builds from settings (evidence: `preprocessing.py`)
- Preprocessing → ECOD calculation: calculates ECOD (evidence: `preprocessing.py`)
- Climate dataset → Configuration: loads settings (evidence: `dataloader.py`)
- Climate dataset → Preprocessing: transforms inputs (evidence: `dataloader.py`)
- Climate dataset → Climate utilities: uses data helpers (evidence: `dataloader.py`)
- Training runner → Configuration: loads settings (evidence: `train.py`)
- Training runner → Climate dataset: loads batches (evidence: `train.py`)
- Training runner → Neural model: trains model (evidence: `train.py`)
- Climate utilities → Configuration: uses variable config (evidence: `utils.py`)
- Climate utilities → Preprocessing: uses transforms (evidence: `utils.py`)
- Climate utilities → Neural model: loads model (evidence: `utils.py`)
- Model evaluation → Configuration: loads settings (evidence: `eval.py`)
- Model evaluation → Preprocessing: transforms inputs (evidence: `eval.py`)
- Model evaluation → Neural model: runs predictions (evidence: `eval.py`)
- Model evaluation → Map utilities: plots maps (evidence: `eval.py`)
- Ensemble evaluation → Configuration: loads settings (evidence: `eval_ensemble.py`)
- Ensemble evaluation → Preprocessing: transforms inputs (evidence: `eval_ensemble.py`)
- Ensemble evaluation → Neural model: runs ensemble (evidence: `eval_ensemble.py`)
- Closure analysis → Configuration: loads settings (evidence: `eval_closure.py`)
- Closure analysis → Map utilities: plots maps (evidence: `eval_closure.py`)
- Ensemble closure → Configuration: loads settings (evidence: `eval_closure_ensemble.py`)
- Model comparisons → Configuration: reads settings (evidence: `make_comparison_tables.py`)
- Model comparisons → Model evaluation: uses evaluation results (evidence: `make_comparison_tables.py`)
- Model comparisons → Preprocess runner: uses preprocessing results (evidence: `make_comparison_tables.py`)
- Autoencoder evaluation → Configuration: loads settings (evidence: `eval_autoencoder.py`)
- Interactive autoencoder → Autoencoder evaluation: uses evaluation (evidence: `eval_autoencoder_interactive.py`)
- Cloud evaluation → Configuration: loads settings (evidence: `eval_cloud.py`)
- Legacy closure → Legacy model evaluation: uses legacy evaluation (evidence: `eval_closure_old_model.py`)
- Legacy closure → Closure analysis: uses closure analysis (evidence: `eval_closure_old_model.py`)
- Legacy model evaluation → Climate utilities: uses climate helpers (evidence: `eval_old_model.py`)
- Shakirova analysis → Configuration: loads settings (evidence: `eval_shakirova.py`)
- Shakirova analysis → Map utilities: plots maps (evidence: `eval_shakirova.py`)
