jeudi 7 avril 2016

IBM Quest Dataset Generator ouputs everything in one line

I am trying to create "Datasets for Frequent Itemset mining / Association Rule Mining" using quest as shown in this site: http://ift.tt/22fBGaD.

One example is this file http://ift.tt/1MVIsk9

This file has 20 items per transaction/line (10,000 lines) .

To do this using quest I need to select the "lit" option:

 seq_data_generator lit -help
 -ntrans number_of_transactions_in_000s (default: 1000)
 -tlen avg_items_per_transaction (default: 10)
 -nitems number_of_different_items_in_000s) (default: 100)

 -npats number_of_patterns (default: 10000)
 -patlen avg_length_of_maximal_pattern (default: 4)
 -corr correlation_between_patterns (default: 0.25)
 -conf avg_confidence_in_a_rule (default: 0.75)

 -fname <filename> (write to filename.data and filename.pat)
 -ascii (default: False)
 -randseed # (reset seed used generate to x-acts; must be negative)
 -version (to print out version info)

If I run this: seq_data_generator lit -ascii -ntrans 10 -tlen 20 -fname c20d10k

I get everything in a single line!! why? how can I use quest to get the same format as the example above??

Aucun commentaire:

Enregistrer un commentaire