11.6.7. astroML.stats.fit_bivariate_normal¶
-
astroML.stats.fit_bivariate_normal(x, y, robust=False)[source]¶ Fit bivariate normal parameters to a 2D distribution of points
- Parameters
- x, yarray_like
The x, y coordinates of the points
- robustboolean (optional, default=False)
If True, then use rank-based statistics which are robust to outliers Otherwise, use mean/std statistics which are not robust
- Returns
- mutuple
(x, y) location of the best-fit bivariate normal
- sigma_1, sigma_2float
The best-fit gaussian widths in the uncorrelated frame
- alphafloat
The rotation angle in radians of the uncorrelated frame