nd-ball / py-irt

Bayesian IRT models in Python
MIT License
118 stars 44 forks source link

4PL Basic Model (no guessing) #5

Closed EntilZha closed 3 years ago

EntilZha commented 3 years ago

Implement initial 4PL IRT model, it has feasibility, but not guessing yet.

Unit test could be improved later, but I_think creating a common test across all models is a better way to do that.

codecov-commenter commented 3 years ago

Codecov Report

Merging #5 (3af039f) into master (9566e80) will increase coverage by 2.00%. The diff coverage is 47.58%.

Impacted file tree graph

@@            Coverage Diff             @@
##           master       #5      +/-   ##
==========================================
+ Coverage   40.88%   42.88%   +2.00%     
==========================================
  Files           7        9       +2     
  Lines         340      485     +145     
==========================================
+ Hits          139      208      +69     
- Misses        201      277      +76     
Impacted Files Coverage Δ
py_irt/models/four_param_logistic.py 26.21% <26.21%> (ø)
tests/test_fourPL.py 100.00% <100.00%> (ø)

Continue to review full report at Codecov.

Legend - Click here to learn more Δ = absolute <relative> (impact), ø = not affected, ? = missing data Powered by Codecov. Last update 9566e80...3af039f. Read the comment docs.

EntilZha commented 3 years ago

Think the issue you pointed out is now fixed

jplalor commented 3 years ago

Yep looks like it. Thanks!