Covariance and Contra-variance w.r.t. row column and inverse change of basis matrix

20 Views Asked by At

Could someone explain in an intuitive and if possible as simple as possible manner (ideally with at least 1 example) the relationship between covariance/contra-variance and row/col vectors ? E.g. like one can multiply e.g. a row vector by a matrix (of compatible size of course) and obtain the same result (but transposed) as multiplying the same matrix with the same vector but on the right, with that vector as a column vector this time (juste the same vector but transposed). It this also related to the contra-variance/covariance concept and if yes how? In math notation: $Ax = (x^TA^T)^T$

Also: is there a relationship with the fact that we have to multiply (a vector in standard basis) by the inverse change-of-basis-matrix to get the coordinates in the new basis, since co/contra-variance seems related to change of basis?

If have read multiple articles (wiki etc) , and understood some aspects but still not found a simple and minimal yet including the ascpects I mentioned, explanation.

Also if possible, I would like a more intuitive (in the sense 'what is the concrete purpose of?' ) covectors... I seems to be like a kind of scalar product but still I cannot understand the meaning, usage and purpose of it and all the "dual space" thing/framework.

(I am aware that there are semi-similar questions on this forum related to this subject but the answers provided to them are not satisfactory for me and I also added the relationship aspects towards inverse change-of-basis-matrix matrix.)