Skip to main content

​
Overview

This module handles the extraction of power plant generation data from OFEI text files. The data contains information about agents (power companies), plant names, types, and 24-hour time series generation data.

​
Data Source Format

The OFEI files follow a specific text format:
  • Agent information marked with AGENTE: prefix
  • Plant records containing type D with comma-separated values
  • 24 hourly generation values per plant

​
Parsing Implementation

import pandas as pd

# Ruta del archivo
ruta = "/content/drive/MyDrive/Prueba_tecnica/Datos3/OFEI1204.txt"

# Lista donde guardaremos los registros finales
data = []

# Variable para almacenar el agente actual
agente_actual = None

with open(ruta, "r", encoding="latin-1") as file:
    for line in file:
        line = line.strip()

        # Detectar líneas de agente
        if line.startswith("AGENTE:"):
            agente_actual = line.replace("AGENTE:", "").strip()

        # Procesar solo líneas que contienen ", D,"
        elif ", D," in line:
            partes = line.split(",")
            planta = partes[0].strip()
            tipo = partes[1].strip()

            # Extraer las 24 horas
            horas = [float(h.strip()) for h in partes[2:26]]

            # Construir registro
            registro = [agente_actual, tipo, planta] + horas

            data.append(registro)

# Crear nombres de columnas
columnas = ["Agente", "Tipo", "Planta"] + [f"Hora_{i}" for i in range(1, 25)]

# Crear DataFrame final
df_final = pd.DataFrame(data, columns=columnas)

​
Parsing Logic

​
Agent Detection

The parser maintains state by tracking the current agent: 
if line.startswith("AGENTE:"):
    agente_actual = line.replace("AGENTE:", "").strip()

​
Plant Record Extraction

Records with type D are parsed: 
elif ", D," in line:
    partes = line.split(",")
    planta = partes[0].strip()
    tipo = partes[1].strip()
    
    # Extraer las 24 horas
    horas = [float(h.strip()) for h in partes[2:26]]

​
Record Structure

Each record combines:
  • Agent name (from previous AGENTE: line)
  • Type (D = generation type)
  • Plant name
  • 24 hourly values (Hora_1 through Hora_24)

​
Output DataFrame Structure

ColumnDescriptionType
AgentePower company namestring
TipoPlant type (D)string
PlantaPlant identifierstring
Hora_1 to Hora_24Hourly generation values (MW)float

​
Sample Output

         Agente Tipo      Planta  Hora_1  Hora_2  ...  Hora_24
0    AES CHIVOR    D     CHIVOR1  125.00  125.00  ...   125.00
1    AES CHIVOR    D     CHIVOR2  125.00  125.00  ...   125.00
2       EMGESA    D     BETANIA   364.00  364.00  ...   364.00
The resulting dataset contains 305 rows representing all active power plants with their 24-hour generation profiles.

​
File Encoding

Note the use of latin-1 encoding to handle special characters in Colombian power plant names: 
with open(ruta, "r", encoding="latin-1") as file:

Build docs developers (and LLMs) love

Get started for freeTalk to us