irt-test 0.1.1

Package for calc difficult of test tasks and ability of test subjects by Rasch model (IRT 0).

irt_test

Package for calc difficult of test tasks and ability of test subjects by Rasch model (IRT 0).

Installation

You can install irt_test using pip:

pip install irt-test

Example Usage

Input DataFrame example:

	Task 1	Task 2	Task 3	Task 4	Task 5	Task 6	Task 7	Task 8	Task 9	Task 10	Task 11
Julia	1	1	1	1	1	1	1	1	0	1	1
Ivan	1	1	1	1	1	1	1	0	1	0	1
Anna	1	1	1	1	1	1	0	1	0	0	1
Peter	1	1	1	1	1	1	0	1	0	0	1
Helen	1	1	1	1	1	1	0	1	0	0	1
Bill	1	1	1	1	1	0	1	0	0	0	1
Tony	1	1	1	1	0	1	0	0	0	0	1
Dmitry	1	0	1	0	1	0	0	0	0	0	1
Natasha	0	1	0	1	0	0	0	0	0	0	1
Jimmy	0	0	0	0	0	0	0	0	0	0	1

For get logits of tasks and subjects use 'irt' function:

from irt_test.irt import irt

# Get logits and additional info from IRT0 model
irt_result = irt(df)

# Abilites of test subjects
irt_result.abilities

	0
Julia	4.72094
Ivan	3.29019
Anna	1.97836
Peter	1.97836
Helen	1.97836
Bill	0.711974
Tony	-0.451242
Dmitry	-2.2367
Natasha	-3.0347

# Difficult of test tasks
irt_result.difficult

	0
Task 1	-2.84779
Task 2	-2.84779
Task 3	-2.84779
Task 4	-2.84779
Task 5	-1.4046
Task 6	-0.212051
Task 7	2.3964
Task 8	1.59305
Task 9	4.50918
Task 10	4.50918

# "IRT model error
irt_result.err

0.0196

# Can't get logits for these subjects
irt_result.rejected_subjects

['Jimmy']

# Can't get logits for these tasks
irt_result.rejected_tasks

['Task 11']