TEXTTOPIC_TRAIN dataset: is it possible to construct it in PROC HPTMINE

[eserates]

I would like to utilize proc hptmine for my text analytics. However as I went through the manual I could not find any property that can help me create the same type of table text miner does. That is TEXT_TOPICTRAIN data. it has TextTopic_raw, TextTopic, _DOCUMENT_, TEXT, ID columns in it.

Can anyone help me with creating the same type of dataset in PROC HPTMINE?

Below is the code I used  but don't see any table there to join to construct a table like texttopic_train data set.

thanks

proc hptmine data=samplestrata20K;

doc_id ID;

var TEXT;

parse

termwgt  = ENTROPY

cellwgt  = log

reducef  = 0

entities = std

synonym  = sashelp.engsynms

stop     = sashelp.engstop

outparent = outparent

outterms  = outterms

outchild  = outchild

outconfig = outconfig;

svd

        k = 25

       outtopics                =  topics

outdocpro      = docpro

TOPICLABELSIZE = 10;

performance details;

run;

[ChrisNZ]

Moved post to SAS Text and Content Analytics

[RussAlbright]

Unfortunately, It isn't possible to use PROC hptmine to get the full topic results found in Text Miner. There is quite a lot of sas code in Text Miner that executes after proc hptmine to do these calculations. If you submit  "options mprint;" in your start up code, you will see it. 

Our newer action on viya, the tmMine action, does have the full computation contained within the action so if you move to that you will direct access to the computation there.

Russ

[eserates]

thank you