Troubleshoot my XGBoost R code, the log model is working

Troubleshoot my XGBoost R code classifier 1 or 0, the logistic model is working against the data set. However, my XGBoost R code does not work against the data set as the confusion matrix produces pred 0 1 0 0 0 1 423 465 when it should produce something like pred 0 1 0 338 350 1 85 115 Deliverables working XGBoost code versus my provided data set broken into training and test subsets with the confusion matrix outcome shown greater than 0.55. Example from the logistics code: [login to view URL] <- predict(m1,type = "response") [login to view URL] <- ifelse([login to view URL] > 0.55, "1", "0") attach(nfl_train) table([login to view URL],ATS) sum([login to view URL] == 1 & ATS == 1 & push == 0)/sum([login to view URL] == 1 & push == 0)