xy-parser
    Preparing search index...

    xy-parser

    xy-parser

    Parse a CSV / TSV / whitespace-separated text file containing scientific data — a spectrum, diffractogram, chromatogram, two-column dump from an instrument, etc. — and return numeric x and y arrays.

    This is the lightweight building block: if you only need {x, y} from a text file, use this. If you also need the technique-aware Analysis wrapper (variables, units, transmittance ↔ absorbance, peak picking), reach for common-spectrum or a domain wrapper like uv-spectrum / ir-spectrum / xrd-analysis, all of which use xy-parser under the hood. For known structured formats use the dedicated parser instead (fromJcamp for JCAMP-DX, spc-parser for Thermo Galactic SPC, …).

    Zakodium logo

    Maintained by Zakodium

    NPM version Test coverage npm download DOI

    npm install xy-parser
    
    import { parseXY, parseXYAndKeepInfo } from 'xy-parser';

    const data = `My file
    1 2
    3 4
    5 6
    7 8`;

    const result = parseXY(data);
    /* result ->
    {
    x: [1, 3, 5, 7],
    y: [2, 4, 6, 8],
    }
    */

    const resultWithInfo = parseXYAndKeepInfo(data);
    /* resultWithInfo ->
    {
    info: [{ position: 0, value: 'My file' }],
    data: {
    x: [1, 3, 5, 7],
    y: [2, 4, 6, 8],
    },
    }
    */

    Columns may be separated by tabs, spaces, commas or semicolons. When columns are separated by tabs, spaces or semicolons, a comma inside a number is read as the decimal separator (1,5\t2,3 gives x: [1.5], y: [2.3]).

    Option Default Description
    xColumn 0 Index of the column used for x.
    yColumn 1 Index of the column used for y.
    bestGuess false Try to determine which columns should be used (see below).
    numberColumns Number.MAX_SAFE_INTEGER Reflow rows into this many columns (e.g. 2 for an x1 y1 x2 y2 … layout).
    minNumberColumns 2 Skip lines with fewer columns.
    maxNumberColumns Number.MAX_SAFE_INTEGER Skip lines with more columns.
    uniqueX false Merge repeated x values, summing their y values.
    rescale false Divide y by its maximum so that the largest value is 1.

    The bestGuess option will try to determine which columns should be used.

    If there are 3 columns and the first column is a sequential number starting at '1' it looks like this is a line number, we will ignore it.

    If there are many columns maybe we have a format like X1, Y1, X2, Y2, ... in this cases if one column on two is a monotone series we will parse it correctly.

    MIT