I'm working on a project and we are using XGBoost to make predictions. My colleague sent me the model file but when I load on my computer it don't run as expected.
When I changed one variable from the model from 0 to 1 it didn't changed the result (in 200 different lines), so I started to investigate. We compared a lot of different results and it was all different.
I run the
xgb_tree it showed the
max_depth is 0, but it is supposed to be 4. When I run
xgb_tree$results it says the
max_depth = 4.
We also tried a lot of different save methods (.rda, .rds, .model) but none of them worked.
Any suggestion would be welcome, thanks.
EDIT: Posting the sessionInfo()
R version 3.2.5 (2016-04-14) Platform: x86_64-w64-mingw32/x64 (64-bit) Running under: Windows >= 8 x64 (build 9200) locale:  LC_COLLATE=English_United States.1252 LC_CTYPE=English_United States.1252  LC_MONETARY=English_United States.1252 LC_NUMERIC=C  LC_TIME=English_United States.1252 attached base packages:  stats graphics grDevices utils datasets methods base other attached packages:  plyr_1.8.4 car_2.1-4 acepack_1.4.1 Ckmeans.1d.dp_3.4.6-4  Hmisc_4.0-0 Formula_1.2-1 survival_2.40-1 memisc_0.99.7-1  MASS_7.3-45 Information_0.0.9 minerva_1.4.5 randomForest_4.6-12  pROC_1.8 xgboost_0.4-4 caret_6.0-73 lattice_0.20-33  ggplot2_2.2.0 miscTools_0.6-22 reshape_0.8.6 data.table_1.9.8  dplyr_0.5.0 e1071_1.6-7 lubridate_1.6.0 psych_1.6.9  readr_1.0.0 stringr_1.1.0 stringi_1.1.2 loaded via a namespace (and not attached):  Rcpp_0.12.8 class_7.3-14 assertthat_0.1 digest_0.6.10  foreach_1.4.3 R6_2.2.0 MatrixModels_0.4-1 stats4_3.2.5  lazyeval_0.2.0 minqa_1.2.4 SparseM_1.74 nloptr_1.0.4  rpart_4.1-10 Matrix_1.2-4 labeling_0.3 splines_3.2.5  lme4_1.1-12 foreign_0.8-66 munsell_0.4.3 compiler_3.2.5  mnormt_1.5-5 mgcv_1.8-12 htmltools_0.3.5 nnet_7.3-12  tibble_1.2 gridExtra_2.2.1 htmlTable_1.7 codetools_0.2-14  ModelMetrics_1.1.0 grid_3.2.5 nlme_3.1-125 gtable_0.2.0  DBI_0.5-1 magrittr_1.5 scales_0.4.1 reshape2_1.4.2  doParallel_1.0.10 latticeExtra_0.6-28 RColorBrewer_1.1-2 iterators_1.0.8  tools_3.2.5 parallel_3.2.5 pbkrtest_0.4-6 colorspace_1.3-1  cluster_2.0.3 knitr_1.15.1 quantreg_5.29
R version 3.3.1 (2016-06-21) Platform: x86_64-w64-mingw32/x64 (64-bit) Running under: Windows >= 8 x64 (build 9200) locale:  LC_COLLATE=Portuguese_Brazil.1252 LC_CTYPE=Portuguese_Brazil.1252 LC_MONETARY=Portuguese_Brazil.1252  LC_NUMERIC=C LC_TIME=Portuguese_Brazil.1252 attached base packages:  stats graphics grDevices utils datasets methods base other attached packages:  caret_6.0-73 ggplot2_2.1.0 lattice_0.20-33 plyr_1.8.4 xgboost_0.4-4  shinydashboard_0.5.3 shiny_0.14.2 loaded via a namespace (and not attached):  Rcpp_0.12.8 nloptr_1.0.4 iterators_1.0.8 tools_3.3.1 digest_0.6.10 lme4_1.1-12  jsonlite_1.1 nlme_3.1-128 gtable_0.2.0 mgcv_1.8-12 Matrix_1.2-6 foreach_1.4.3  parallel_3.3.1 SparseM_1.72 stringr_1.1.0 MatrixModels_0.4-1 stats4_3.3.1 grid_3.3.1  nnet_7.3-12 data.table_1.9.6 R6_2.1.3 minqa_1.2.4 reshape2_1.4.2 car_2.1-3  magrittr_1.5 scales_0.4.0 codetools_0.2-14 ModelMetrics_1.1.0 htmltools_0.3.5 MASS_7.3-45  splines_3.3.1 rsconnect_0.4.3 pbkrtest_0.4-6 mime_0.5 xtable_1.8-2 colorspace_1.2-6  httpuv_1.3.3 quantreg_5.29 stringi_1.1.1 munsell_0.4.3 chron_2.3-47