pad — Pad strings, character arrays, and cell arrays to a target length using MATLAB-compatible options.

pad adds characters to the beginning, end, or both sides of strings so that each element reaches a specified length. It mirrors MATLAB semantics for string arrays, character arrays, and cell arrays of character vectors, including direction keywords, default space padding, and optional custom characters.

How pad works

• Without a target length, pad extends each element to match the longest text in the input.
• Providing a numeric target guarantees a minimum length; existing text that already meets or exceeds the target is returned unchanged.
• Direction keywords ('left', 'right', 'both') are case-insensitive; 'right' is the default. When an odd number of pad characters is required for 'both', the extra character is appended to the end.
• padChar must be a single character (string scalar or 1×1 char array). The default is a space.
• Character arrays remain rectangular. Each row is padded independently and then widened with spaces so the array keeps MATLAB’s column-major layout.
• Cell arrays preserve their structure. Elements must be string scalars or 1×N character vectors and are padded while keeping their original type.
• Missing strings (string(missing)) and empty character vectors pass through unchanged, preserving metadata.

How RunMat runs pad on the GPU

pad always executes on the CPU. When an argument (or a value nested inside a cell array) lives on the GPU, RunMat gathers it, performs the padding step, and produces a host result or re-wraps the padded value inside the cell. No provider hooks exist yet for string padding, so providers and fusion planners treat pad as a sink that terminates device residency.

GPU memory and residency

No. Text data in RunMat lives on the host today. If text happens to originate from a GPU computation, pad automatically gathers it before padding, so you never have to manage residency manually for this builtin.

Examples

Pad Strings To A Common Width

labels = ["GPU"; "Accelerate"; "RunMat"];
aligned = pad(labels)

Expected output:

aligned =
  3×1 string
    "GPU       "
    "Accelerate"
    "RunMat    "

Pad Strings On The Left With Zeros

ids = ["42"; "7"; "512"];
zero_padded = pad(ids, 4, 'left', '0')

Expected output:

zero_padded =
  3×1 string
    "0042"
    "0007"
    "0512"

Center Text With Both-Sided Padding

titles = ["core"; "planner"];
centered = pad(titles, 10, 'both', '*')

Expected output:

centered =
  2×1 string
    "***core***"
    "*planner**"

Pad Character Array Rows

chars = char("GPU", "RunMat");
out = pad(chars, 8)

Expected output:

out =

  2×8 char array

    'GPU     '
    'RunMat  '

Pad A Cell Array Of Character Vectors

C = {'solver', "planner", 'jit'};
cell_out = pad(C, 'right', '.')

Expected output:

cell_out = 1×3 cell array
    {'solver.'}    {"planner"}    {'jit....'}

Leave Missing Strings Unchanged

values = ["RunMat", "<missing>", "GPU"];
kept = pad(values, 8)

Expected output:

kept =
  1×3 string
    "RunMat  "    <missing>    "GPU     "

FAQ

Related Strings functions

Open-source implementation

Unlike proprietary runtimes, every RunMat function is open-source. Read exactly how pad works, line by line, in Rust.

• View pad.rs on GitHub
• Learn how the runtime works
• Found a bug? Open an issue with a minimal reproduction.

About RunMat

RunMat is an open-source runtime that executes MATLAB-syntax code — faster, on any GPU, with no license required.

• Simulations that took hours now take minutes. RunMat automatically optimizes your math for GPU execution on Apple, Nvidia, and AMD hardware. No code changes needed.
• Start running code in seconds. Open the browser sandbox or download a single binary. No license server, no IT ticket, no setup.
• A full development environment. GPU-accelerated 2D and 3D plotting, automatic versioning on every save, and a browser IDE you can share with a link.

Getting started · Benchmarks · Pricing