Skip to content

Fix fitting (for logistic regression) - #74

Draft
enryH wants to merge 36 commits into
mainfrom
fix_fitting
Draft

Fix fitting (for logistic regression)#74
enryH wants to merge 36 commits into
mainfrom
fix_fitting

Conversation

@enryH

@enryH enryH commented Jul 1, 2026

Copy link
Copy Markdown
Collaborator

  • add examples for model fitting

    • base it on mechanistic model and plot the fitted parameters in phenomological

  • fitting functions take linear scale measurements

    • adapt for phenomological models the fit to log-transform the data
    • update inference to use N0 as min(N) and update inference to correctly generate the data

enryH added 30 commits July 1, 2026 13:58
@enryH
📝 update docstrings of models and developing docs
2146bee 
- highlight equations
- distinguish between lag and intial value of measurement curve
@enryH
📝 update equation and docstrings, leave hints
5058a66 
- group fitting functions explicitly
@enryH
🎨 group helper functions for mechanistic and phenomological models se…
182f97a 
…parately
@enryH
🚧 change fitting and adapt inference to new fitting (plus add example…
85b9e07 
… to docs)
@enryH
🚚 prepare to move fitting function to own subpackage
d5da0cc 
- fit.py is equivalent as fit/__init__.py w.r.t to importing it
@enryH
🐛 add missing import (forgot to commit)
53dd5f5
@enryH
🎨 change toc (adapt headings)
53d679b
@enryH
🐛 fix fitting phenomological models
53b5270 
- realize that ln(N0) has to be fitted as function itself is non-zero at t=0 in most cases
@enryH
🚧 test sympy and create more plots
f86625d
@enryH
🎨 remove imports from within functions
0640818
@enryH
🐛 let time-span at least start from zero
f44aa90 
- negative times are not tested. should they exist?
@enryH
🎨 format files
a673554
@enryH
🚧 more plots
fc2d71e
@enryH
🎨 update docstrings of phenom. models
4e2b73d
@enryH
🎨 add interactive figure to explain how to compute core metric and wh…
a25fabb 
…at N0 is

- N0 = N(t=lag)
- N(0) is non-zero itself

Both leads to some additional sublte complications to calculate a global shift in y
@enryH
🐛 try to render equation of Sphinx
76c4e2a
@enryH
🎨 format
c454d97
@enryH
🚧 test if equations can be shown on RTD
3b96a0f 
https://jupyter-book.readthedocs.io/v1/content/math.html
@enryH
🔥 remove old config
9be37de
@enryH
🚧 try to put it on one line
304a0f1
@enryH
🎨 test displaying eq
2405f9e
@enryH
⏪ go back to only display relevant formula
dea3df5
@enryH
🎨 update description of logistic models, reorder equations
bdd3d24
@enryH
🚧 see if any combination makes formulas visisble in ipynb
3f419ab
@enryH
✨ recover mu and mu_max using mech_logistic model
262b353
@enryH
🎨 plot similar looking growth curves (using classic and paper model)
a714a8c
@enryH
🚧 update examples for logistic regression
6e5a7e0
@enryH
🐛 allow to fit negative lag (in case data was recoreded too late)
4f03a2b 
- ln_N used to allow for an offset to be calculated (needs to be documented for fitting)
@enryH
🎨 fit non-unique solution of logistic model in linear space (to recov…
63694a2 
…er mu)

- lag estimation is arbitrary -> only use heuristics for lag estimation.
@enryH
🎨 add non-parametric fitting of synthetic data
f06729e 
- key point: OD_phen_classic has no real lack in log-space (as it is a linear)
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

No reviews

Assignees

No one assigned

Labels

None yet

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

1 participant

@enryH