Re: How to implement 'zero-order constraint' or free network adjustment

[Matt Newville]

On Fri, Sep 2, 2022 at 9:20 AM phisan...@gmail.com <phisan...@gmail.com> wrote:

Hello, 

i am geodesist and trying to implement optimizing of a levelling network. 
Goal is to minimize the observed height differences connected into
close loops. As you might refer to Dr.Deakin ...here..
http://www.mygeodesy.id.au/documents/Free%20Net%20Level%20AdjustmentV2.pdf
     

The question is :-
1 the covariance matrix is look different

I don't understand what that means.  Is that a question?  I don't see a covariance matrix at all.

2 all residuals via LMFIT is just close to Dr.Deaking's calculation.

Is that a question?   Would that assume that the reader would have read and understood the linked PDF?

3 For sake of network free constraint.... we have to define 
      A+B+X+Y+Z = 0.
    i do adding --> params.add( name="A", expr= '-(B+X+Y+Z)' )
    Is there any other design ?

Probably not.   Is that approach causing problems?

--Matt