第100+16步 ChatGPT学习：R实现Xgboost分类

基于R 4.2.2版本演示

一、写在前面

有不少大佬问做机器学习分类能不能用R语言，不想学Python咯。

答曰：可！用GPT或者Kimi转一下就得了呗。

加上最近也没啥内容写了，就帮各位搬运一下吧。

二、R代码实现Xgboost分类

（1）导入数据

我习惯用RStudio自带的导入功能：

（2）建立Xgboost模型（默认参数）

# Load necessary libraries
library(caret)
library(pROC)
library(ggplot2)
library(xgboost)

# Assume 'data' is your dataframe containing the data
# Set seed to ensure reproducibility
set.seed(123)

# Split data into training and validation sets (80% training, 20% validation)
trainIndex <- createDataPartition(data$X, p = 0.8, list = FALSE)
trainData <- data[trainIndex, ]
validData <- data[-trainIndex, ]

# Prepare matrices for XGBoost
dtrain <- xgb.DMatrix(data = as.matrix(trainData[, -which(names(trainData) == "X")]), label = trainData$X)
dvalid <- xgb.DMatrix(data = as.matrix(validData[, -which(names(validData) == "X")]), label = validData$X)

# Define parameters for XGBoost
params <- list(booster = "gbtree", 
               objective = "binary:logistic", 
               eta = 0.1, 
               gamma = 0, 
               max_depth = 6, 
               min_child_weight = 1, 
               subsample = 0.8, 
               colsample_bytree = 0.8)

# Train the XGBoost model
model <- xgb.train(params = params, data = dtrain, nrounds = 100, watchlist = list(eval = dtrain), verbose = 1)

# Predict on the training and validation sets
trainPredict <- predict(model, dtrain)
validPredict <- predict(model, dvalid)

# Convert predictions to binary using 0.5 as threshold
#trainPredict <- ifelse(trainPredict > 0.5, 1, 0)
#validPredict <- ifelse(validPredict > 0.5, 1, 0)

# Calculate ROC curves and AUC values
#trainRoc <- roc(response = trainData$X, predictor = as.numeric(trainPredict))
#validRoc <- roc(response = validData$X, predictor = as.numeric(validPredict))
trainRoc <- roc(response = as.numeric(trainData$X) - 1, predictor = trainPredict)
validRoc <- roc(response = as.numeric(validData$X) - 1, predictor = validPredict)

# Plot ROC curves with AUC values
ggplot(data = data.frame(fpr = trainRoc$specificities, tpr = trainRoc$sensitivities), aes(x = 1 - fpr, y = tpr)) +
  geom_line(color = "blue") +
  geom_area(alpha = 0.2, fill = "blue") +
  geom_abline(slope = 1, intercept = 0, linetype = "dashed", color = "black") +
  ggtitle("Training ROC Curve") +
  xlab("False Positive Rate") +
  ylab("True Positive Rate") +
  annotate("text", x = 0.5, y = 0.1, label = paste("Training AUC =", round(auc(trainRoc), 2)), hjust = 0.5, color = "blue")

ggplot(data = data.frame(fpr = validRoc$specificities, tpr = validRoc$sensitivities), aes(x = 1 - fpr, y = tpr)) +
  geom_line(color = "red") +
  geom_area(alpha = 0.2, fill = "red") +
  geom_abline(slope = 1, intercept = 0, linetype = "dashed", color = "black") +
  ggtitle("Validation ROC Curve") +
  xlab("False Positive Rate") +
  ylab("True Positive Rate") +
  annotate("text", x = 0.5, y = 0.2, label = paste("Validation AUC =", round(auc(validRoc), 2)), hjust = 0.5, color = "red")

# Calculate confusion matrices based on 0.5 cutoff for probability
confMatTrain <- table(trainData$X, trainPredict >= 0.5)
confMatValid <- table(validData$X, validPredict >= 0.5)

# Function to plot confusion matrix using ggplot2
plot_confusion_matrix <- function(conf_mat, dataset_name) {
  conf_mat_df <- as.data.frame(as.table(conf_mat))
  colnames(conf_mat_df) <- c("Actual", "Predicted", "Freq")
  
  p <- ggplot(data = conf_mat_df, aes(x = Predicted, y = Actual, fill = Freq)) +
    geom_tile(color = "white") +
    geom_text(aes(label = Freq), vjust = 1.5, color = "black", size = 5) +
    scale_fill_gradient(low = "white", high = "steelblue") +
    labs(title = paste("Confusion Matrix -", dataset_name, "Set"), x = "Predicted Class", y = "Actual Class") +
    theme_minimal() +
    theme(axis.text.x = element_text(angle = 45, hjust = 1), plot.title = element_text(hjust = 0.5))
  
  print(p)
}

# Now call the function to plot and display the confusion matrices
plot_confusion_matrix(confMatTrain, "Training")
plot_confusion_matrix(confMatValid, "Validation")

# Extract values for calculations
a_train <- confMatTrain[1, 1]
b_train <- confMatTrain[1, 2]
c_train <- confMatTrain[2, 1]
d_train <- confMatTrain[2, 2]

a_valid <- confMatValid[1, 1]
b_valid <- confMatValid[1, 2]
c_valid <- confMatValid[2, 1]
d_valid <- confMatValid[2, 2]

# Training Set Metrics
acc_train <- (a_train + d_train) / sum(confMatTrain)
error_rate_train <- 1 - acc_train
sen_train <- d_train / (d_train + c_train)
sep_train <- a_train / (a_train + b_train)
precision_train <- d_train / (b_train + d_train)
F1_train <- (2 * precision_train * sen_train) / (precision_train + sen_train)
MCC_train <- (d_train * a_train - b_train * c_train) / sqrt((d_train + b_train) * (d_train + c_train) * (a_train + b_train) * (a_train + c_train))
auc_train <- roc(response = trainData$X, predictor = trainPredict)$auc

# Validation Set Metrics
acc_valid <- (a_valid + d_valid) / sum(confMatValid)
error_rate_valid <- 1 - acc_valid
sen_valid <- d_valid / (d_valid + c_valid)
sep_valid <- a_valid / (a_valid + b_valid)
precision_valid <- d_valid / (b_valid + d_valid)
F1_valid <- (2 * precision_valid * sen_valid) / (precision_valid + sen_valid)
MCC_valid <- (d_valid * a_valid - b_valid * c_valid) / sqrt((d_valid + b_valid) * (d_valid + c_valid) * (a_valid + b_valid) * (a_valid + c_valid))
auc_valid <- roc(response = validData$X, predictor = validPredict)$auc

# Print Metrics
cat("Training Metrics\n")
cat("Accuracy:", acc_train, "\n")
cat("Error Rate:", error_rate_train, "\n")
cat("Sensitivity:", sen_train, "\n")
cat("Specificity:", sep_train, "\n")
cat("Precision:", precision_train, "\n")
cat("F1 Score:", F1_train, "\n")
cat("MCC:", MCC_train, "\n")
cat("AUC:", auc_train, "\n\n")

cat("Validation Metrics\n")
cat("Accuracy:", acc_valid, "\n")
cat("Error Rate:", error_rate_valid, "\n")
cat("Sensitivity:", sen_valid, "\n")
cat("Specificity:", sep_valid, "\n")
cat("Precision:", precision_valid, "\n")
cat("F1 Score:", F1_valid, "\n")
cat("MCC:", MCC_valid, "\n")
cat("AUC:", auc_valid, "\n")

在R语言中，训练Xgboost模型时，可调参数很多：

1）通用参数

这些参数用于控制XGBoost的整体功能：

①booster: 选择每一步的模型类型，常用的有：

1. gbtree：基于树的模型（默认）
2. gblinear：线性模型
3. dart：Dropouts meet Multiple Additive Regression Trees

②nthread: 并行线程数，默认为最大可用线程数。

③verbosity: 打印消息的详细程度，0 (silent), 1 (warning), 2 (info), 3 (debug)。

2）Booster 参数：

控制每一步提升（booster）的行为：

①eta (或 learning_rate): 学习率，控制每步的收缩以防止过拟合。

②min_child_weight: 决定最小叶子节点样本权重和，用于控制过拟合。

③max_depth: 树的最大深度，限制树的增长以避免过拟合。

④max_leaf_nodes: 最大叶子节点数。

⑤gamma (或 min_split_loss): 分裂节点所需的最小损失函数下降值。

⑥subsample: 训练每棵树时用于随机采样的部分数据比例。

⑦colsample_bytree/colsample_bylevel/colsample_bynode: 构建树时每个级别的特征采样比例。

⑧lambda (或 reg_lambda): L2 正则化项权重。

⑨alpha (或 reg_alpha): L1 正则化项权重。

⑩scale_pos_weight: 在类别不平衡的情况下加权正类的权重。

n_estimators / nrounds：Boosting 过程中的树的数量，或者说是提升迭代的轮数。每轮迭代通常会产生一个新的模型（通常是一棵树）。

3）学习任务参数

用于控制学习任务和相应的学习目标：

①objective: 定义学习任务和相应的学习目标，如：

②binary:logistic: 二分类的逻辑回归，返回预测概率。

③multi:softmax: 多分类的softmax，需要设置 num_class（类别数）。

④reg:squarederror: 回归任务的平方误差。

⑤eval_metric: 验证数据的评估指标，如 rmse、mae、logloss、error (分类错误率)、auc 等。

⑥seed: 随机数种子，用于结果的可重复性。

5）DART Booster特有参数

当 booster 设置为 dart 时：

①sample_type: 采样类型。

②normalize_type: 归一化类型。

③rate_drop: 每次迭代中树的丢弃率。

④skip_drop: 跳过丢弃的概率。

在随便设置了一些参数值，结果如下：

从AUC来看，Xgboost随便一跑直接就过拟合了，验证集的性能相比训练集差距挺大的。得好好调参调参才行。

三、Xgboost手动调参原则

调参的一般策略是，可以先使用网格搜索（Grid Search）、随机搜索（Random Search）或更高级的方法如贝叶斯优化来粗略地确定合适的参数范围，然后在这些范围内细致地调整和验证，以找到最优的模型配置。

主要调的参数：max_depth、min_child_weight、gamma、subsample、colsample_bytree / colsample_bylevel / colsample_bynode、eta、lambda、alpha和n_estimators (或 nrounds)。

max_depth（最大深度）：通常范围是3到10。较大的深度可能会导致过拟合，尤其是在小数据集上。

min_child_weight（最小子节点权重）：有助于控制过拟合。面对高度不平衡的类别时，可以适当增加此值。

gamma（伽马）：从0开始调整，根据控制过拟合的需要逐渐增加。

subsample、colsample_bytree/colsample_bylevel/colsample_bynode（子采样率、按树/层/节点的列采样率）：通常范围从0.5到1。这些参数控制了每一步的数据子采样。

eta（学习率）：较小的值可以使训练更加稳健，但需要更多的训练迭代。

lambda 和 alpha（L2和L1正则化项）：在成本函数中添加正则化项。0到10的范围通常效果不错。

nrounds（树的数量，或迭代次数）：更多的树可以模拟更复杂的模式，但也可能导致过拟合。

# Load necessary libraries
library(caret)
library(pROC)
library(ggplot2)
library(xgboost)

# Assume 'data' is your dataframe containing the data
# Set seed to ensure reproducibility
set.seed(123)

# Convert the target variable to factor if not already
data$X <- factor(data$X)

# Split data into training and validation sets (80% training, 20% validation)
trainIndex <- createDataPartition(data$X, p = 0.8, list = FALSE)
trainData <- data[trainIndex, ]
validData <- data[-trainIndex, ]

# Prepare matrices for XGBoost
dtrain <- xgb.DMatrix(data = as.matrix(trainData[, -which(names(trainData) == "X")]), label = as.numeric(trainData$X) - 1)
dvalid <- xgb.DMatrix(data = as.matrix(validData[, -which(names(validData) == "X")]), label = as.numeric(validData$X) - 1)

# Define parameter grid
depths <- c(4, 6, 10)
weights <- c(1, 5, 10)
gammas <- c(0, 0.2, 0.5)
subsamples <- c(0.5, 0.8, 0.9)
colsamples <- c(0.5, 0.8, 0.9)
etas <- c(0.01, 0.1, 0.2)
lambdas <- c(0, 5, 10)
alphas <- c(0, 1, 5)
nrounds <- c(100, 250, 500)

best_auc <- 0
best_params <- list()

# Loop through parameter grid
for (max_depth in depths) {
  for (min_child_weight in weights) {
    for (gamma in gammas) {
      for (subsample in subsamples) {
        for (colsample_bytree in colsamples) {
          for (eta in etas) {
            for (lambda in lambdas) {
              for (alpha in alphas) {
                for (nround in nrounds) {
                  # Set parameters for this iteration
                  params <- list(
                    booster = "gbtree",
                    objective = "binary:logistic",
                    eta = eta,
                    gamma = gamma,
                    max_depth = max_depth,
                    min_child_weight = min_child_weight,
                    subsample = subsample,
                    colsample_bytree = colsample_bytree,
                    lambda = lambda,
                    alpha = alpha
                  )
                  
                  # Train the model
                  model <- xgb.train(params = params, data = dtrain, nrounds = nround, watchlist = list(eval = dtrain), verbose = 0)
                  
                  # Predict on the validation set
                  validPredict <- predict(model, dvalid)
                  validPredictBinary <- ifelse(validPredict > 0.5, 1, 0)
                  
                  # Calculate AUC
                  validRoc <- roc(response = as.numeric(validData$X) - 1, predictor = validPredictBinary)
                  auc_score <- auc(validRoc)
                  
                  # Update best model if current AUC is better
                  if (auc_score > best_auc) {
                    best_auc <- auc_score
                    best_params <- params
                    best_params$nrounds <- nround
                  }
                }
              }
            }
          }
        }
      }
    }
  }
}

# Print the best AUC and corresponding parameters
print(paste("Best AUC:", best_auc))
print("Best Parameters:")
print(best_params)

# After parameter tuning, train the model with best parameters
model <- xgb.train(params = best_params, data = dtrain, nrounds = best_params$nrounds, watchlist = list(eval = dtrain), verbose = 0)

# Predict on the training and validation sets using the final model
trainPredict <- predict(model, dtrain)
validPredict <- predict(model, dvalid)

# Convert predictions to binary using 0.5 as threshold
#trainPredictBinary <- ifelse(trainPredict > 0.5, 1, 0)
#validPredictBinary <- ifelse(validPredict > 0.5, 1, 0)

# Calculate ROC curves and AUC values
#trainRoc <- roc(response = trainData$X, predictor = as.numeric(trainPredict))
#validRoc <- roc(response = validData$X, predictor = as.numeric(validPredict))
trainRoc <- roc(response = as.numeric(trainData$X) - 1, predictor = trainPredict)
validRoc <- roc(response = as.numeric(validData$X) - 1, predictor = validPredict)

# Plot ROC curves with AUC values
ggplot(data = data.frame(fpr = trainRoc$specificities, tpr = trainRoc$sensitivities), aes(x = 1 - fpr, y = tpr)) +
  geom_line(color = "blue") +
  geom_area(alpha = 0.2, fill = "blue") +
  geom_abline(slope = 1, intercept = 0, linetype = "dashed", color = "black") +
  ggtitle("Training ROC Curve") +
  xlab("False Positive Rate") +
  ylab("True Positive Rate") +
  annotate("text", x = 0.5, y = 0.1, label = paste("Training AUC =", round(auc(trainRoc), 2)), hjust = 0.5, color = "blue")

ggplot(data = data.frame(fpr = validRoc$specificities, tpr = validRoc$sensitivities), aes(x = 1 - fpr, y = tpr)) +
  geom_line(color = "red") +
  geom_area(alpha = 0.2, fill = "red") +
  geom_abline(slope = 1, intercept = 0, linetype = "dashed", color = "black") +
  ggtitle("Validation ROC Curve") +
  xlab("False Positive Rate") +
  ylab("True Positive Rate") +
  annotate("text", x = 0.5, y = 0.2, label = paste("Validation AUC =", round(auc(validRoc), 2)), hjust = 0.5, color = "red")

# Calculate confusion matrices based on 0.5 cutoff for probability
confMatTrain <- table(trainData$X, trainPredict >= 0.5)
confMatValid <- table(validData$X, validPredict >= 0.5)

# Function to plot confusion matrix using ggplot2
plot_confusion_matrix <- function(conf_mat, dataset_name) {
  conf_mat_df <- as.data.frame(as.table(conf_mat))
  colnames(conf_mat_df) <- c("Actual", "Predicted", "Freq")
  
  p <- ggplot(data = conf_mat_df, aes(x = Predicted, y = Actual, fill = Freq)) +
    geom_tile(color = "white") +
    geom_text(aes(label = Freq), vjust = 1.5, color = "black", size = 5) +
    scale_fill_gradient(low = "white", high = "steelblue") +
    labs(title = paste("Confusion Matrix -", dataset_name, "Set"), x = "Predicted Class", y = "Actual Class") +
    theme_minimal() +
    theme(axis.text.x = element_text(angle = 45, hjust = 1), plot.title = element_text(hjust = 0.5))
  
  print(p)
}

# Now call the function to plot and display the confusion matrices
plot_confusion_matrix(confMatTrain, "Training")
plot_confusion_matrix(confMatValid, "Validation")

# Extract values for calculations
a_train <- confMatTrain[1, 1]
b_train <- confMatTrain[1, 2]
c_train <- confMatTrain[2, 1]
d_train <- confMatTrain[2, 2]

a_valid <- confMatValid[1, 1]
b_valid <- confMatValid[1, 2]
c_valid <- confMatValid[2, 1]
d_valid <- confMatValid[2, 2]

# Training Set Metrics
acc_train <- (a_train + d_train) / sum(confMatTrain)
error_rate_train <- 1 - acc_train
sen_train <- d_train / (d_train + c_train)
sep_train <- a_train / (a_train + b_train)
precision_train <- d_train / (b_train + d_train)
F1_train <- (2 * precision_train * sen_train) / (precision_train + sen_train)
MCC_train <- (d_train * a_train - b_train * c_train) / sqrt((d_train + b_train) * (d_train + c_train) * (a_train + b_train) * (a_train + c_train))
auc_train <- roc(response = trainData$X, predictor = trainPredict)$auc

# Validation Set Metrics
acc_valid <- (a_valid + d_valid) / sum(confMatValid)
error_rate_valid <- 1 - acc_valid
sen_valid <- d_valid / (d_valid + c_valid)
sep_valid <- a_valid / (a_valid + b_valid)
precision_valid <- d_valid / (b_valid + d_valid)
F1_valid <- (2 * precision_valid * sen_valid) / (precision_valid + sen_valid)
MCC_valid <- (d_valid * a_valid - b_valid * c_valid) / sqrt((d_valid + b_valid) * (d_valid + c_valid) * (a_valid + b_valid) * (a_valid + c_valid))
auc_valid <- roc(response = validData$X, predictor = validPredict)$auc

# Print Metrics
cat("Training Metrics\n")
cat("Accuracy:", acc_train, "\n")
cat("Error Rate:", error_rate_train, "\n")
cat("Sensitivity:", sen_train, "\n")
cat("Specificity:", sep_train, "\n")
cat("Precision:", precision_train, "\n")
cat("F1 Score:", F1_train, "\n")
cat("MCC:", MCC_train, "\n")
cat("AUC:", auc_train, "\n\n")

cat("Validation Metrics\n")
cat("Accuracy:", acc_valid, "\n")
cat("Error Rate:", error_rate_valid, "\n")
cat("Sensitivity:", sen_valid, "\n")
cat("Specificity:", sep_valid, "\n")
cat("Precision:", precision_valid, "\n")
cat("F1 Score:", F1_valid, "\n")
cat("MCC:", MCC_valid, "\n")
cat("AUC:", auc_valid, "\n")

结果输出：

以上是找到的相对最优参数组合，看看具体性能：

似乎有点提升，过拟合没那么明显了。验证集的性能也有所提高。

有兴趣可以继续调参。

五、最后

数据嘛：

链接：https://pan.baidu.com/s/1rEf6JZyzA1ia5exoq5OF7g?pwd=x8xm

提取码：x8xm