LSTM PV prediction

Quick start

1. Install dependencies:

pip install -r requirements.txt

2. Train (example):

python src/models/train_lstm_pv.py --data examples/LSTM3miso/completeDataF.csv --epochs 30

3. Predict next value:

python src/models/predict_lstm_pv.py --model models/lstm_pv.h5 --scaler models/scaler.pkl --data examples/LSTM3miso/completeDataF.csv

Notes

• The scripts follow the preprocessing from examples/LSTM3miso/misopredict.ipynb (default uses columns 2:8 and time_steps=24). Adjust arguments if your CSV differs.
• Model and scaler are saved under models/ by default.

agents

Weather Agent con Mock Server

El Weather Agent puede funcionar con o sin conexión a internet:

# Modo normal (intenta NASA API, fallback a mock)
python src/agents/weather.py

# Modo simulación (siempre usa mock)
./launch_mock_weather.sh  # Terminal 1: Servidor mock
python src/agents/weather.py --mock  # Terminal 2: Agente

Ver documentación completa del Mock Server

Otros Agentes

python src/agents/solar.py
python src/agents/weather.py --mock 
python src/agents/dashboard.py 
python tools/mock_weather_server.py 

pip install flask