Data for: Ensemble species distribution modelling with transformed suitability values

The data set provides results from 10 runs (ensemble 1 - 10) of calibrating an ensemble species distribution model with two different methods of selecting cross-validation data sets (random, block) and two different methods of transforming suitability values (no transformation, probit transformation). AUC_1 to AUC_4 show the AUC for each 4-fold crossvalidation tests. AUC_MIN and AUC_MEAN are the minimum and average of AUC_1 to AUC_4. The Thryothorus data set gives the results for Thryothorus ludovicianus (Case study 2). The Bradypus data set gives the results for Bradypus variegatus (Case study 1). Point location data sets for both species and bioclimatic raster layers are available from the dismo and maxlike R packages as documented in the manuscript and the R script available from the Appendix.

本数据集包含10次运行（集成集1至10）的校准结果，该运行针对集成物种分布模型（ensemble species distribution model）开展校准，校准过程采用了两种不同的交叉验证数据集选择方法（随机法、区块法），以及两种不同的适宜性数值转换方式（无转换、概率单位变换（probit transformation））。AUC_1至AUC_4分别对应4次4折交叉验证测试的受试者工作特征曲线下面积（AUC）；AUC_MIN与AUC_MEAN则为AUC_1至AUC_4的最小值与平均值。Thryothorus数据集包含路易斯安那苇鹪鹩（Thryothorus ludovicianus）的实验结果（案例研究2）；Bradypus数据集包含褐喉树懒（Bradypus variegatus）的实验结果（案例研究1）。两种物种的点位数据集与生物气候栅格图层均可从dismo和maxlike R包中获取，相关说明已在论文手稿及附录提供的R脚本中详述。