🤖 This is an automated PR from Repo Assist.

Closes #444

Summary

Implements iloc-style integer-position based indexing for both Series and Frame, equivalent to the pandas iloc accessor.

New API

Series.iloc

// Select elements at positions 0, 2, 4
let s = series [ "a" => 10; "b" => 20; "c" => 30; "d" => 40; "e" => 50 ]
s |> Series.iloc [0; 2; 4]
// → series [ "a" => 10; "c" => 30; "e" => 50 ]

Original keys are preserved.

Frame.ilocRows

// Select rows at positions 0 and 2
df |> Frame.ilocRows [0; 2]

Frame.ilocCols

// Select columns at positions 0 and 2
df |> Frame.ilocCols [0; 2]

Frame.iloc

// Select rows 1, 2, 3 and columns 1, 2 (equivalent to pandas iloc[1:4, 1:3])
df |> Frame.iloc [1; 2; 3] [1; 2]

Implementation

All four functions resolve integer positions to the underlying keys using the existing GetRowKeyAt / ColumnIndex.KeyAt methods, then delegate to frame.Rows.[keys] / frame.Columns.[keys] / series.GetItems(keys). This keeps the implementation minimal and leverages battle-tested code paths.

The functions are also compiled under PascalCase names (Iloc, IlocRows, IlocCols) so they are usable from C#.

Test Status

✅ 7 new tests added in Frame.fs and Series.fs
✅ All 595 tests pass (dotnet test tests/Deedle.Tests/Deedle.Tests.fsproj -c Release)

Generated by Repo Assist for issue #444 · ◷

To install this agentic workflow, run

gh aw add githubnext/agentics/workflows/repo-assist.md@30f2254f2a7a944da1224df45d181a3f8faefd0d