Using joined RDFs: FilterAvailable error on defined column that uses an aliased column and FilterAvailable error on an aliased column.

[modregominguez]

Check duplicate issues.

• Checked for duplicates

Description

After using RDataFrame Alias on a coulmn and then Define a new column using this aliased column, if one uses FilterAvailable on either the aliased column or the new defined column errors appear.

For TTree case:

• FilterAvailable on the DEFINED column:
  RDataFrame: Define could not retrieve value for column 'met_met_NOSYS' for entry 1. You can use the DefaultValueFor operation to provide a default value, or FilterAvailable/FilterMissing to discard/keep entries with missing values instead.
• FilterAvailable on the Alias column:
  The resulting RDF reports 0 entries.

For the RNTuple case:

• FilterAvailable on the DEFINED column:
  RDataFrame: Define could not retrieve value for column 'reco.met_met_NOSYS' for entry 1. You can use the DefaultValueFor operation to provide a default value, or FilterAvailable/FilterMissing to discard/keep entries with missing values instead.
• FilterAvailable on the Alias column:
  RNTupleProcessorDS: Column "matched_field" does not exist

Reproducer

Here I attach the reproducer. To see the different errors I just uncomment the section I want to test.
I also attach the root files required for testing.

Here it is the link to the CERN box with the files: https://cernbox.cern.ch/s/vmbIyGbFs8VFUpv

ROOT version

For the ttree part the root versions used to find the same errors were:
ROOT Version: 6.40.00
Built for linuxx8664gcc on May 26 2026, 15:56:51
From tags/v6-40-00@v6-40-00

ROOT Version: 6.41.01
Built for linuxx8664gcc on Jun 15 2026, 08:27:46
From heads/ntuple-ds-processor@v6-39-99-585-g56d04c01

For the RNTuple part the root version used was:
ROOT Version: 6.41.01
Built for linuxx8664gcc on Jun 15 2026, 08:27:46
From heads/ntuple-ds-processor@v6-39-99-585-g56d04c01

Installation method

Build from source

Operating system

Ubuntu 25.10

Additional context

No response

[vepadulano]

Dear @modregominguez,

Thank you for the report and the concise reproducer! This indeed looks like a bug, I will take a look at it.