Input
- glmpredict_admissions, created as follows:
CREATE MULTISET TABLE glmpredict_admissions AS ( SELECT * FROM GLMPredict_MLE ( ON admissions_test PARTITION BY ANY ON glm_admissions_model AS Model DIMENSION USING Accumulate ('id', 'masters', 'gpa', 'stats', 'programming', 'admitted') Family ('LOGISTIC') LinkFunction ('LOGIT') OutputProb ('t') Responses ('1') ) AS dt ) WITH DATA;
The preceding call is the same as in GLMPredict_MLE Example: Logistic Distribution Prediction, except for the addition of Responses ('1').
SELECT * FROM glmpredict_admissions;
id masters gpa stats programming admitted fitted_value prediction prob_1 ----------- --------------------- -------- ----------- -------- ---------------------- ----------- ---------------------- 61 yes 4.00000000000000E 000 Advanced Advanced 1 6.50620999077464E-001 1 6.50620999077464E-001 51 yes 3.75999999046326E 000 Beginner Beginner 0 3.55711267427044E-001 0 3.55711267427044E-001 57 no 3.71000003814697E 000 Advanced Advanced 1 9.46412427155178E-001 1 9.46412427155178E-001 59 no 3.65000009536743E 000 Novice Novice 1 8.74190793835473E-001 1 8.74190793835473E-001 60 no 4.00000000000000E 000 Advanced Novice 1 8.65060169870818E-001 1 8.65060169870818E-001 68 no 1.87000000476837E 000 Advanced Novice 1 8.90966498700309E-001 1 8.90966498700309E-001 55 no 3.59999990463257E 000 Beginner Advanced 1 9.68031454705773E-001 1 9.68031454705773E-001 53 yes 3.50000000000000E 000 Beginner Novice 1 5.56015243694066E-001 1 5.56015243694066E-001 58 no 3.13000011444092E 000 Advanced Advanced 1 9.49666668893439E-001 1 9.49666668893439E-001 66 no 3.86999988555908E 000 Novice Beginner 1 7.54740372192837E-001 1 7.54740372192837E-001 62 no 3.70000004768372E 000 Advanced Advanced 1 9.46470180931532E-001 1 9.46470180931532E-001 69 no 3.96000003814697E 000 Advanced Advanced 1 9.44949341902683E-001 1 9.44949341902683E-001 56 no 3.81999993324280E 000 Advanced Advanced 1 9.45773244683836E-001 1 9.45773244683836E-001 64 yes 3.80999994277954E 000 Advanced Advanced 1 6.55525616200381E-001 1 6.55525616200381E-001 54 yes 3.50000000000000E 000 Beginner Advanced 1 7.69476126601993E-001 1 7.69476126601993E-001 65 yes 3.90000009536743E 000 Advanced Advanced 1 6.53206426068888E-001 1 6.53206426068888E-001 52 no 3.70000004768372E 000 Novice Beginner 1 7.58307989347034E-001 1 7.58307989347034E-001 63 no 3.82999992370605E 000 Advanced Advanced 1 9.45714782104350E-001 1 9.45714782104350E-001 50 yes 3.95000004768372E 000 Beginner Beginner 0 3.50765681699298E-001 0 3.50765681699298E-001 67 yes 3.46000003814697E 000 Novice Beginner 0 2.60036217847503E-001 0 2.60036217847503E-001
SQL Call
SELECT * FROM ROC ( ON glmpredict_admissions AS InputTable OUT TABLE ROCTable (roc_out_4) USING ProbabilityColumn ('prob_1') ObservationColumn ('admitted') PositiveClass ('1') NumThresholds (100) ) AS dt;
Output
Onscreen:
model_id auc gini -------- ---------------------- ---------------------- 1 1.00000000000000E 000 1.00000000000000E 000
SELECT * FROM roc_out_4;
model_id threshold tpr fpr -------- ---------------------- ---------------------- ---------------------- 1 0.00000000000000E 000 1.00000000000000E 000 1.00000000000000E 000 1 1.01010101010101E-002 1.00000000000000E 000 1.00000000000000E 000 1 2.02020202020202E-002 1.00000000000000E 000 1.00000000000000E 000 ... 1 9.09090909090909E-002 1.00000000000000E 000 1.00000000000000E 000 1 1.01010101010101E-001 1.00000000000000E 000 1.00000000000000E 000 1 1.11111111111111E-001 1.00000000000000E 000 1.00000000000000E 000 1 1.21212121212121E-001 1.00000000000000E 000 1.00000000000000E 000 ... 1 1.91919191919192E-001 1.00000000000000E 000 1.00000000000000E 000 1 2.02020202020202E-001 1.00000000000000E 000 1.00000000000000E 000 1 2.12121212121212E-001 1.00000000000000E 000 1.00000000000000E 000 1 2.22222222222222E-001 1.00000000000000E 000 1.00000000000000E 000 ... 1 2.92929292929293E-001 1.00000000000000E 000 6.66666666666667E-001 1 3.03030303030303E-001 1.00000000000000E 000 6.66666666666667E-001 1 3.13131313131313E-001 1.00000000000000E 000 6.66666666666667E-001 1 3.23232323232323E-001 1.00000000000000E 000 6.66666666666667E-001 ... 1 3.93939393939394E-001 1.00000000000000E 000 0.00000000000000E 000 1 4.04040404040404E-001 1.00000000000000E 000 0.00000000000000E 000 1 4.14141414141414E-001 1.00000000000000E 000 0.00000000000000E 000 1 4.24242424242424E-001 1.00000000000000E 000 0.00000000000000E 000 ... 1 4.94949494949495E-001 1.00000000000000E 000 0.00000000000000E 000 1 5.05050505050505E-001 1.00000000000000E 000 0.00000000000000E 000 1 5.15151515151515E-001 1.00000000000000E 000 0.00000000000000E 000 1 5.25252525252525E-001 1.00000000000000E 000 0.00000000000000E 000 ... 1 5.95959595959596E-001 9.41176470588235E-001 0.00000000000000E 000 1 6.06060606060606E-001 9.41176470588235E-001 0.00000000000000E 000 1 6.16161616161616E-001 9.41176470588235E-001 0.00000000000000E 000 1 6.26262626262626E-001 9.41176470588235E-001 0.00000000000000E 000 ... 1 6.96969696969697E-001 7.64705882352941E-001 0.00000000000000E 000 1 7.07070707070707E-001 7.64705882352941E-001 0.00000000000000E 000 1 7.17171717171717E-001 7.64705882352941E-001 0.00000000000000E 000 1 7.27272727272727E-001 7.64705882352941E-001 0.00000000000000E 000 ... 1 7.97979797979798E-001 5.88235294117647E-001 0.00000000000000E 000 1 8.08080808080808E-001 5.88235294117647E-001 0.00000000000000E 000 1 8.18181818181818E-001 5.88235294117647E-001 0.00000000000000E 000 1 8.28282828282828E-001 5.88235294117647E-001 0.00000000000000E 000 ... 1 8.98989898989899E-001 4.11764705882353E-001 0.00000000000000E 000 1 9.09090909090909E-001 4.11764705882353E-001 0.00000000000000E 000 1 9.19191919191919E-001 4.11764705882353E-001 0.00000000000000E 000 1 9.29292929292929E-001 4.11764705882353E-001 0.00000000000000E 000 ... 1 9.89898989898990E-001 0.00000000000000E 000 0.00000000000000E 000 1 1.00000000000000E 000 0.00000000000000E 000 0.00000000000000E 000
Download a zip file of all examples and a SQL script file that creates their input tables.