PCAGaussianFitter

Class for fitting a multidimensional gaussian distribution to a PCA-reduced point cloud

Instead of fitting the full (high-dimensional) point cloud, a PCA-based dimensionality reduction is first applied on it. This has the advantage that the fit can be visualised correctly (in case of 2 reduced dimensions), instead of only projections of it. The potential disadvantage is that the PCA reduction might distort the relative separations.

---

---

[class] PCAGaussianFitter

comments:

class for fitting a gaussian distribution to a PCA-reduced point cloud  
parameters  
- pca: sklearn.decomposition.pca object  
- mean: multidim mean of normal distribution  
- cov: multidim covariance matrix of normal distribution  
- mvn: scipy.stats multivariate_normal object built from mean and cov  

⤷ __init__

full signature:

def __init__(self)  

comments:

empty constructor  
input arguments:  

⤷ fit

full signature:

def fit(self, points, npcadims=2)  

comments:

fit to a set of points  
input arguments:  
- points: a np array of shape (npoints,ndims)  
- npcadims: number of PCA compoments to keep  

⤷ pdf

full signature:

def pdf(self, points)  

comments:

get pdf at points  
note: points can be both of shape (npoints,ndims) or of shape (npoints,npcadims);  
      in the latter case it is assumed that the points are already PCA-transformed,  
      and only the gaussian kernel density is applied on them.  

⤷ transform

full signature:

def transform(self, points)  

comments:

perform PCA transformation  

---