TD_RandomProjectionFit Input Table
Each of the input data points in the following input table consists of 963 columns and a unique identifier of the data point:
company_id date_2010_01_04 date_2010_01_05 .... date_2013_10_28 date_2013_10_29 1 0.58 -0.220005 0.840019 -19.589981 2 -0.640002 -0.65 -0.400002 0.66 3 -2.350006 1.260009 .... -1.760009 3.740021 4 0.109997 0 0.040001 0.540001 5 0.459999 1.77 .... 1.130005 0.309998 6 0.45 0.460001 -0.06 -0.11 7 0.18 0.220001 .... 0.330002 1.150001 8 0.73 0.369999 0.090001 -0.110001 9 0.899997 0.700001 .... -0.220001 0.159996 10 0.36 0.909996 1.070003 1.050003
TD_RandomProjectionFit SQL Call
SELECT * FROM TD_RandomProjectionFit(
ON stock_movement as InputTable
OUT PERMANENT TABLE OutputTable(rand_proj_fit_tbl_ex)
USING
TargetColumns('[1:]')
NumComponents(353)
Seed(0)
Epsilon(0.25)
ProjectionMethod('GAUSSIAN')
) as dt ORDER BY 1,2;
Output Table
td_rpj_feature_353_0 date_2010_01_04 date_2010_01_05 .... date_2013_10_28 date_2013_10_29 0 -0.068021509 -0.021542237 .... 0.031534748 -0.021109446 1 0.076606803 0.033381927 .... 0.047328448 0.039908753 2 -0.100981365 0.026625478 .... -0.035816093 -0.009469693 .... .... .... .... .... .... .... .... .... .... .... .... 351 0.039755885 -0.040469189 .... 0.046338113 0.018347439 352 0.040278453 0.072748211 .... -0.04456492 -0.031728421