De-Grouping a table

KNIME Analytics Platform
1

Hi,

does somebody know about a node or trick that can do the following transformation:

3 | 1 | 0
0 | 0 | 1

into

1 | 1 | 0
0 | 0 | 1
1 | 0 | 0
1 | 0 | 0

i need a transformation like theres no value above 1 . I need this for the bitvector-generator for association rule mining.

the problem is that i get the same rules with the same confidence and support for the following tables:

3 | 1 | 0
0 | 0 | 1

1 | 1 | 0
0 | 0 | 1

and this is wrong i think. It would be very nice if someone have a tip for me about that.

greetings

User

2

does noone got an idea for me?

I think when i directly apply the bitvector generator on

3 | 1 | 0
0 | 0 | 1

i will get wrong result. is that correct? because the result on

1 | 1 | 0
0 | 0 | 1

should be different as the on first table. and they are equal.

greetings User

3

Can you explain your input? And your configuration in the Bitvector generator?

My understanding would be, there are 3 items A, 1 items B and no items C in the shopping basket.
so, the corresponding bit vector would be 110 same as there are 1 item A, 1 item B and no items C.

The assoziation rule only cares for the combination of products, not for the frequency.

4

ok here we go:

after pivoting and so on i get (for example) the following table:
|customer | A | B | C|
1001 |1 | 1 | 0 |
1002 |0 | 0 | 1 |

(A,B,C are the products)
with that table there no problem to use the bitvector generator.
In the bitvector-generator i use the option ‘numeric input’ and threshold 1

but what do i do if i get the table

|customer | A | B | C|
1001 |3 | 1 | 0 |
1002 |0 | 0 | 1 |

if i directly apply the bitvector generator i will loose some information. if i rise up the threshold to 3 i wont get assoc rules.

My questtion is:
What can i do to get correct assoc rules. ???

greetings
User