Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Handling correlations between variables #4

Open
cgobat opened this issue Jan 13, 2023 · 0 comments
Open

Handling correlations between variables #4

cgobat opened this issue Jan 13, 2023 · 0 comments
Labels
enhancement New feature or request help wanted Extra attention is needed

Comments

@cgobat
Copy link
Owner

cgobat commented Jan 13, 2023

This package currently has very limited capacity for correctly handling correlations between variables. For example, for two quantities with uncertainties $x\pm\sigma_x$ and $y\pm\sigma_y$, $(x-x) \not\equiv (x-y)$, even if $x=y$ and $\sigma_x = \sigma_y$. This is because $x-x\equiv0\pm0$, but $x-y=0\pm\sigma_{x-y}$, with $\sigma_{x-y}\neq0$.

The functionality to handle this specific subtraction case is already planned (i.e., a_u.__sub__(self, other) checks if other is self), but in general the propagation formulae implemented by this package assume uncorrelated operands. The uncertainties package for symmetric error propagation handles this correctly, and it would be nice if this package could too. Checks for identity should be fairly straightforward to implement, but users may also wish to indicate correlation between two variables that don't share an address in memory (i.e., self is other is False). Some other way of keeping track of (auto)corrrelation may therefore eventually be called for/warranted.

@cgobat cgobat added enhancement New feature or request help wanted Extra attention is needed labels Jan 13, 2023
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
enhancement New feature or request help wanted Extra attention is needed
Projects
None yet
Development

No branches or pull requests

1 participant