Detección de fraude con Pandas: caso con registros empresariales – Operating systems, scripting, PowerShell and security

Python

Crear el fichero CSV con registros empresariales (hay una columna de fraude)

import pandas as pd
import numpy as np

# Seed for reproducibility
np.random.seed(42)

# Number of records
num_records = 100

# Generate sample data
data = {
    'transaction_id': np.arange(1, num_records + 1),
    'company_name': [f'Empresa_{i}' for i in np.random.randint(1, 20, num_records)],
    'transaction_amount': np.round(np.random.uniform(50, 5000, num_records), 2),
    'transaction_date': pd.to_datetime('2023-01-01') + pd.to_timedelta(np.random.randint(0, 365, num_records), unit='D'),
    'employee_id': np.random.randint(1000, 1050, num_records),
    'department': np.random.choice(['Ventas', 'Marketing', 'Finanzas', 'RRHH', 'IT'], num_records)
}

df = pd.DataFrame(data)

# Introduce some fraudulent transactions (e.g., higher amounts, specific employee/company combos)
# Let's say 10% of transactions are fraudulent
fraud_indices = np.random.choice(df.index, size=int(num_records * 0.1), replace=False)
df['is_fraud'] = 0
df.loc[fraud_indices, 'is_fraud'] = 1

# Make some fraudulent transactions have unusually high amounts
df.loc[df['is_fraud'] == 1, 'transaction_amount'] = np.round(np.random.uniform(7000, 20000, df['is_fraud'].sum()), 2)

# Save the DataFrame to a CSV file
csv_file_path = 'registros_empresa.csv'
df.to_csv(csv_file_path, index=False)

print(f"CSV file '{csv_file_path}' created successfully with {num_records} records.")
print(f"Number of fraudulent records: {df['is_fraud'].sum()}")

import pandas as pd

import numpy as np

# Seed for reproducibility

np.random.seed(42)

# Number of records

num_records = 100

# Generate sample data

data = {

'transaction_id': np.arange(1, num_records + 1),

'company_name': [f'Empresa_{i}' for i in np.random.randint(1, 20, num_records)],

'transaction_amount': np.round(np.random.uniform(50, 5000, num_records), 2),

'transaction_date': pd.to_datetime('2023-01-01') + pd.to_timedelta(np.random.randint(0, 365, num_records), unit='D'),

'employee_id': np.random.randint(1000, 1050, num_records),

'department': np.random.choice(['Ventas', 'Marketing', 'Finanzas', 'RRHH', 'IT'], num_records)

}

df = pd.DataFrame(data)

# Introduce some fraudulent transactions (e.g., higher amounts, specific employee/company combos)

# Let's say 10% of transactions are fraudulent

fraud_indices = np.random.choice(df.index, size=int(num_records * 0.1), replace=False)

df['is_fraud'] = 0

df.loc[fraud_indices, 'is_fraud'] = 1

# Make some fraudulent transactions have unusually high amounts

df.loc[df['is_fraud'] == 1, 'transaction_amount'] = np.round(np.random.uniform(7000, 20000, df['is_fraud'].sum()), 2)

# Save the DataFrame to a CSV file

csv_file_path = 'registros_empresa.csv'

df.to_csv(csv_file_path, index=False)

print(f"CSV file '{csv_file_path}' created successfully with {num_records} records.")

print(f"Number of fraudulent records: {df['is_fraud'].sum()}")

Cargar el fichero CSV con Pandas

import pandas as pd

# Load the CSV file into a pandas DataFrame
csv_file_path = 'registros_empresa.csv'
df_loaded = pd.read_csv(csv_file_path)

# Display the first 5 rows to verify
print("Primeras 5 filas del DataFrame cargado:")
display(df_loaded.head())

# Display basic information
print("\nInformación del DataFrame cargado:")
display(df_loaded.info())

import pandas as pd

# Load the CSV file into a pandas DataFrame

csv_file_path = 'registros_empresa.csv'

df_loaded = pd.read_csv(csv_file_path)

# Display the first 5 rows to verify

print("Primeras 5 filas del DataFrame cargado:")

display(df_loaded.head())

# Display basic information

print("\nInformación del DataFrame cargado:")

display(df_loaded.info())

Detectar valores que tienen 1 en columna de fraude

# Filtrar el DataFrame para mostrar solo las transacciones fraudulentas
fraudulent_transactions = df_loaded[df_loaded['is_fraud'] == 1]

print("Transacciones fraudulentas:\n")
display(fraudulent_transactions)

print(f"\nNúmero total de transacciones fraudulentas: {len(fraudulent_transactions)}")

# Filtrar el DataFrame para mostrar solo las transacciones fraudulentas

fraudulent_transactions = df_loaded[df_loaded['is_fraud'] == 1]

print("Transacciones fraudulentas:\n")

display(fraudulent_transactions)

print(f"\nNúmero total de transacciones fraudulentas: {len(fraudulent_transactions)}")

Detección de fraude con Pandas: caso con registros empresariales (Python)

LIBRO DE POWERSHELL

LO ÚLTIMO

RECOMENDADO