Error DataSpec generated by configure does not match spec after execution

[Jeff]

Hi,

I’m trying to figure out an issue with a workflow. I have a list files node that looks through a directory of xlsx files. Then with another node I read the names of the sheet names within each excel file. Then with excel Reader I read the data from a certain sheet name copying the data to process. Since there is about a thousand files I’m not sure which files are causing problems but 3 times I see the above error.

While reading through the community I see a few articles referring to the error but I don’t see how to correct it.

I’m wondering two things…

1. how one debugs workflows which hopefully result in determining
2. how I correct the error.

Any guidance?

Thanks,
Jeff

[izaychik63]

In Preview press Refresh.

[Jeff]

Thanks for response. Trying to understand your answer…I’m not sure where I press refresh. I don’t see a preview node… so not sure where I should look.

Thanks!
Jeff

[ipazin]

Hi there @Jeff,

which KNIME version are you running? Which node throws this error?

Is your workflow design similar to last one from this workflow example?

Br,
Ivan

[Jeff]

Yes very similar to shown workflow. I am running v4.1 and the node that gives the error is the Excel Reader (XLS) node.

Jeff

[ipazin]

Hi,

I see. But does workflow executes successfully or it stops? Also should data from all sheets and files have same structure or not?

Br,
Ivan

[Jeff]

Yes workflow successfully completes despite the three errors. My hope is that the structure is the same but thinking there may be three files with a different structure so trying to figure out a way to capture which files are causing problems.

Thanks,
Jeff

[mlauber71]

Since you are dealing with the import of Excel files and the type of columns this thread might also be relevant for you

Excel Reader max file size

One way would be to try and use R to import the Excel files and also write the results directly into SQLite, ARFF or Parquet (or CSV). So they would not need to be processed in KNIME (but you could always load them back from the mentioned formats). If the Excel files become more complex you could also use R to help you with that (sheet names) etc. If you have problems with column formats you could force them all to string and then decide the right format If you have to deal with pas…

[ipazin]

Hi @Jeff,

Managed to reproduce it. Probably order of columns is different in those 3 Excel files but according to what I have seen probably doesn’t impact concatenated data as it concatenates according to column name.

Anyways will check this a bit more and get back to you.

Br,
Ivan

[ipazin]

Hi there @Jeff,

just to follow up on this one. Have created a ticket for it and hopefully will end confusion in one of following releases.

Regarding finding out which files have different structure - there are Table Validator and Table Validator (Reference) nodes but they do not catch column order.

Br,
Ivan