polynomial regression with specify options return NaN

[throrin19]

Hi, I've a problem with the polynomial regression if I set the option (empty or not) :

var series 	= [[0, 10], [1 ,3], [2, 0], [3, 2], [4, 4], [5, 0], [6, 0], [7, 0], [8, 1], [9, 3]];

// works fine
var result = regression('polynomial', series);

// does not works
var result = regression('polynomial', series, {});
var result = regression('polynomial', series, { order : 2 });
var result = regression('polynomial', series, { order : 3 });
// ...

For information, I use that with the CDN version of regression (version 1.4.0 : https://cdnjs.cloudflare.com/ajax/libs/regression/1.4.0/regression.min.js)

And this is the result if this does not works :

{
  "r2": null,
  "equation": [
    "[object Object]1"
  ],
  "points": [
    [
      0,
      null
    ],
    [
      1,
      null
    ],
    [
      2,
      null
    ],
    [
      3,
      null
    ],
    [
      4,
      null
    ],
    [
      5,
      null
    ],
    [
      6,
      null
    ],
    [
      7,
      null
    ],
    [
      8,
      null
    ],
    [
      9,
      null
    ]
  ],
  "string": "y = NaN"
}

[ruidiasbraga]

I am also having the same problem? any solution?
This is related with the Math.log(); with zero values, return "-infinity" and translates to "NaN" but I don't want to make a custom code for this, I prefer to wait and update the file from github.

[timoklingenhoefer]

the error occurs here:

const coefficients = gaussianElimination(rhs, k).map((v) => round(v, options.precision));

the polynomial regression needs the precision option. Otherwise the round function tries to devide by null.
So you have to change

var result = regression('polynomial', series, { order : 2 });

to

var result = regression('polynomial', series, { order : 2, precision: 3 });