Tabular: relabel by row counter

Tool

toolshed.g2.bx.psu.edu/repos/bgruening/text_processing/tp_awk_tool. This is the operation-named awk recipe for replacing row content with deterministic labels derived from NR.

When to reach for it

Use this when a workflow needs synthetic row labels like sample_0, sample_1, … or sample_1, sample_2, … based only on input row order.

Do not use this for simple computed columns; use tabular-compute-new-column. Do not use this when labels should come from collection element identifiers; use collection-aware provenance such as tabular-concatenate-collection-to-table.

Parameters

• code: awk program. A one-line quoted string is enough for this operation.
• infile: connected input.
• variables: usually [] in awk examples.

The corpus exemplar uses toolshed.g2.bx.psu.edu/repos/bgruening/text_processing/tp_awk_tool/9.5+galaxy3.

Idiomatic shapes

Corpus-observed sample_0 first row:

tool_id: toolshed.g2.bx.psu.edu/repos/bgruening/text_processing/tp_awk_tool/9.5+galaxy3
tool_state:
  code: '{gsub( $0 ,"sample_" (NR-1)); print}'
  infile: { __class__: ConnectedValue }
  variables: []

Anchored by the MAGs binning evaluation IWC exemplar.

Clearer equivalent for new whole-row replacement:

tool_state:
  code: '{print "sample_" (NR-1)}'
  infile: { __class__: ConnectedValue }
  variables: []

One-based labels:

tool_state:
  code: '{print "sample_" NR}'
  infile: { __class__: ConnectedValue }
  variables: []

Keep a header and relabel only data rows:

tool_state:
  code: 'NR==1 {print; next} {print "sample_" (NR-1)}'
  infile: { __class__: ConnectedValue }
  variables: []

Pitfalls

• Zero-based vs one-based labels. Corpus emits sample_0 first because it uses NR-1. Use NR for sample_1.
• Headers shift counters. Decide whether the first line is preserved, dropped, or counted before picking NR vs NR-1.
• Avoid gsub($0, ...) for new work. It works in the exemplar, but treats the current row as a regex pattern. print "sample_" ... is clearer when replacing the whole row.
• Whole-row replacement destroys original columns. Only use when downstream wants labels only.
• Row order becomes semantic. Any upstream sort or filter changes assigned sample numbers.
• No collection provenance. If labels should come from element identifiers, do not synthesize them from row order.

See also

• iwc-tabular-operations-survey — §2g and §7 awk split decision.
• tabular-prepend-header — header-preserving variants.
• tabular-compute-new-column — preserve rows while adding/replacing columns.
• tabular-concatenate-collection-to-table — preserve collection element identity instead of synthesizing row labels.