A 'total least squares'-like hack for LOESS. Works by rotating points 45 degrees, fitting LOESS, and rotating back.

1 |

`x` |
X values. |

`y` |
Y values. |

`theta` |
Amount to rotate, sets the ratio of variences that are assumed by the hack. Default: -pi/4 radians (45 degrees) for orthogonal regression. |

`span` |
The LOESS span parameter. Default: 1 |

List of input values and LOESS predictions.

Charles G. Danko

Questions? Problems? Suggestions? Tweet to @rdrrHQ or email at ian@mutexlabs.com.

All documentation is copyright its authors; we didn't write any of that.