Unexpected NaN values when training XGBoost model with categorical features
After trying multiple solutions online, I still can't figure this out. I'm working on a project and hit a roadblock... I'm having trouble with I've been banging my head against this for hours. I'm currently working on a machine learning project using XGBoost version 1.5.0 to classify some categorical data. However, I've encountered a baffling scenario where my model returns NaN values during training when I include one-hot encoded features. I've pre-processed my dataset using pandas to one-hot encode categorical variables, but after converting them into DMatrix, the model throws an behavior and stops training. Here’s the code snippet that I’m using: ```python import pandas as pd import xgboost as xgb # Load dataset data = pd.read_csv('data.csv') # One-hot encoding categorical variables data = pd.get_dummies(data, columns=['category_col'], drop_first=True) # Preparing features and labels X = data.drop('target', axis=1) Y = data['target'] # Convert to DMatrix dtrain = xgb.DMatrix(X, label=Y) # Set parameters for XGBoost params = { 'objective': 'binary:logistic', 'eval_metric': 'logloss', } # Train model model = xgb.train(params, dtrain, num_boost_round=100) ``` When I run this code, I get the following behavior message: ``` ValueError: NaN values in the feature matrix. ``` I've double-checked the dataset for any NaN values before encoding and ensured that all columns are numeric after encoding, but the behavior continues. I also tried using `X.fillna(0)` to check if that resolves the scenario, but to no avail. I suspect it might be related to how the one-hot encoding interacts with XGBoost's DMatrix, but I’m unsure how to troubleshoot this further. Can anyone suggest what might be going wrong here or how to properly handle categorical features with XGBoost? Any insights would be greatly appreciated! This is part of a larger application I'm building. I'm working on a application that needs to handle this. Thanks for your help in advance! I'm on Ubuntu 22.04 using the latest version of Python. I'm developing on Ubuntu 20.04 with Python. Could this be a known issue?