npm:transpluck | Skypack

README

transpluck

Reformats csv-like array-of-array with header row into a dataframe-like object in a transpose/pluck-like operation

Installation

npm i transpluck -S

Dependencies

None. Suitable for usage on nodejs or on the browser, via browserify.

Initialization

const transpluck = require('transpluck');

Usage

transpluck(csvData) uses first row as header row, starts decoding at 2nd row, extracts everything

csvData = [ 
    ['a','b','c'],
    [1, 2, 3],
    [2, 7, 1],
    [8, 5, 6]
    ];

seriesData = transpluck(csvData);
// {a: [1,2,8], b: [2,7,5], c: [3,1,6] }

transpluck(csvData, {pluck: ['b']}) returns only the plucked columns, using csvData[0] as the header row. The pluck property must be an array, and should consist of a subset of the column names from the 0th row of the data. Missing items will be ignored, and will not appear as keys in the result.

--> {b: [2,7,5]}

transpluck(csvData, ['x','y','z']) uses the explicit header ['x','y','z'], treats all rows of csvData as data

--> {x: ['a',1,2,8], y: ['b',2,7,5], z: ['c',3,1,6]}

If an explicit header is supplied, a starting row index can also be supplied:

transpluck(csvData, ['x','y','z'], 1)

--> {x: [1,2,8], y: [2,7,5], z: [3,1,6] }

Any explicitly undefined header slots are omitted in the returned object:

transpluck(csvData, ['a',,'c'], 1)

--> {a: [1,2,8], c: [3,1,6]}

You may use an object as the second parameter to request a sparse set of columns. The sparse header object format is {colNumber: 'label', colNumber2: 'label2', ...}.

`transpluck(csvData, {1:'bb'}, 1)

--> {bb: [2,7,5]}

Remember that any header row in the csvData array-of-arrays is ignored when an explicit header is used:

transpluck(csvData, ['a','c'], 1)

--> {a: [1,2,8], c: [2,7,5] } not c:[3,1,6] because explicit header ['a','c'] means collect column 0 in property a, and column 1 in property c.

Ill-formed calls return undefined

transpluck() === undefined

transpluck([1,2,3,4,5,6]) === undefined

transpluck({}) === undefined

Tests

Use mocha framework.

Copyright

License

MIT

Usage no npm install needed!