The following example assumes sparse input format for InputTable:
create multiset table scale_input_sparse (passenger int, attribute_column varchar(20), attribute_value real); insert into scale_input_sparse values (97, 'age', 71); insert into scale_input_sparse values (97, 'fare', 34.6542); insert into scale_input_sparse values (488, 'age', 58); insert into scale_input_sparse values (488, 'fare', 29.7); insert into scale_input_sparse values (505, 'age', 16); insert into scale_input_sparse values (505, 'fare', 86.5); insert into scale_input_sparse values (631, 'age', 80); insert into scale_input_sparse values (631, 'fare', 30); insert into scale_input_sparse values (873, 'age', 33); insert into scale_input_sparse values (873, 'fare', 5);
FitTable Generated with TD_ScaleFit Function
TD_STATTYPE_SCLFIT attribute_column attribute_value
------------------ ---------------- ---------------
avg fare 3.71708400000000E 001
count fare 5.00000000000000E 000
globalscale_false fare NULL
intercept fare 0.00000000000000E 000
location fare 5.00000000000000E 000
max fare 8.65000000000000E 001
min fare 5.00000000000000E 000
missvalue_KEEP fare NULL
multiplier fare 1.00000000000000E 000
null fare 0.00000000000000E 000
scale fare 8.15000000000000E 001
ScaleMethodNumberMapping: [0:mean,1:sum,2:ustd,3:std,4:range,5:midrange,6:maxabs,7:rescale]
fare 4.00000000000000E 000
sum fare 1.85854200000000E 002
TD_ScaleTransform SQL Call
SELECT * FROM TD_ScaleTransform(
ON scale_input_sparse AS InputTable
ON sparseScaleFitOut AS FitTable DIMENSION
USING
AttributeNameColumn('attribute_column')
AttributeValueColumn('attribute_value')
Accumulate('passenger')
)as dt2 order by 1,2;
TD_ScaleTransform Output
passenger attribute_column attribute_value --------- ---------------- --------------- 97 fare 0.363855214723926 488 fare 0.303067484662577 505 fare 1.00000000000000E 000 631 fare 0.306748466257669 873 fare 0.00000000000000E 000