Pivot sees data before a string replace node (bug?)

Hi all,

When working with R, column names have to be carefully chosen to avoid problems with R code. I pre-processed a CSV file using String Replace nodes to replace the data. This is fine. The resulting table looks like this (all rows shown from my workflow using interactive table -> write CSV):
“row ID”,“Sample”,“Gene”,“Per Gene Mean CT Value”
“GAPDH”,“1 do coleoptile”,“GAPDH”,21.58
"GAPDH_0",“2 do coleoptile”,“GAPDH”,21.0775
"GAPDH_1",“3 do coleoptile”,“GAPDH”,20.465
"GAPDH_2",“4 do coleoptile”,“GAPDH”,21.447499999999998
"GAPDH_3",“5 do coleoptile”,“GAPDH”,21.41
"GAPDH_4",“6 do coleoptile”,“GAPDH”,24.235
"GAPDH_5",“7 do coleoptile”,“GAPDH”,23.619999999999997
"a.tubulin",“1 do coleoptile”,“a.tubulin”,21.0775
"a.tubulin_0",“2 do coleoptile”,“a.tubulin”,20.384999999999998
"a.tubulin_1",“3 do coleoptile”,“a.tubulin”,19.4575
"a.tubulin_2",“4 do coleoptile”,“a.tubulin”,19.86
"a.tubulin_3",“5 do coleoptile”,“a.tubulin”,20.015
"a.tubulin_4",“6 do coleoptile”,“a.tubulin”,23.3825
"a.tubulin_5",“7 do coleoptile”,“a.tubulin”,24.1325
"actin",“1 do coleoptile”,“actin”,33.037499999999994
"actin_0",“2 do coleoptile”,“actin”,34.245
"actin_1",“3 do coleoptile”,“actin”,32.6425
"actin_2",“4 do coleoptile”,“actin”,32.114999999999995
"actin_3",“5 do coleoptile”,“actin”,32.665
"actin_4",“6 do coleoptile”,“actin”,35.4825
"actin_5",“7 do coleoptile”,“actin”,33.99
"cyclophilin",“1 do coleoptile”,“cyclophilin”,15.132499999999999
"cyclophilin_0",“2 do coleoptile”,“cyclophilin”,15.01
"cyclophilin_1",“3 do coleoptile”,“cyclophilin”,15.2025
"cyclophilin_2",“4 do coleoptile”,“cyclophilin”,15.684999999999999
"cyclophilin_3",“5 do coleoptile”,“cyclophilin”,15.17
"cyclophilin_4",“6 do coleoptile”,“cyclophilin”,16.5425
"cyclophilin_5",“7 do coleoptile”,“cyclophilin”,16.235

With a-tubulin replaced by a.tubulin for R’s benefit. However, then a pivot is performed and it produces a column with the pre-string replaced data! For brevity, only a few rows are displayed:
“row ID”,“actin”,“a-tubulin”,“cyclophilin”,“GAPDH”,“a.tubulin”
“1 do coleoptile”,33.037499999999994,15.132499999999999,21.58,21.0775
"2 do coleoptile",34.245,15.01,21.0775,20.384999999999998

The pivot node was configured with pivot column set to Gene, group column set to Sample, aggregation enabled (mean method over Per Gene Mean VT Value column), highlighting off, ignore missing values checked.

These results are with knime 2.0.2, running under 32bit Windows XP.

Hi acassin,
Can you please end me the whole flow which demonstrates the actual problem…?
Cheers, Thomas

Hi acassin,

Exactly which column is not the same as the input table you’ve laid out here? Is it the pivoted (column “headers”) in the new dataset?

Best regards,