Added functions to the chi_square test library to

oliver-czulo commented 7 months ago

test contingency tables for independence of observations

estebanz01 commented 7 months ago

thank you! 🙏 I promise I'll review this week. Lots of stuff going on at the same time.

oliver-czulo commented 7 months ago

thank you! 🙏 I promise I'll review this week. Lots of stuff going on at the same time.

No worries, took me long enough how to do a rebase properly 🙈 I grew up with SVN, and GIT concepts can be quite different... Looking forward to your feedback, any time 😃

estebanz01 commented 6 months ago

Hi! Sorry for the looooong review. I was reading lots of theory around it to make sure we have a robust scenario.

Can you add the following information as a test case? After that, we just need to merge it!

Relationship between Region and Product

Alpha: 0.05
Chi-score ~ 38.2484
p-value ~ 0.00000100449

source: https://gplresearch.com/prueba-chi-cuadrado/

Region	Product
	A	B	C
N	27	12	41
S	53	65	29
E	27	32	53
W	12	18	15

Independence between nutrition and IQ

Alpha: 0.1
Chi-score ~ 9.75
Null hypothesis rejected, alternative accepted.

source: https://www.estadistica.net/Algoritmos2/web-contingencia.pdf (Contraste de Independencia Example, page 27).

Nutrition	IQ
	< 80	80 - 90	90 - 99	>= 100
Good	245	228	177	219
Bad	31	27	13	10

Political Affiliation and Opinion

Alpha: 0.05
Chi-score ~ 22.152
p-value <= 0.0001

source: https://online.stat.psu.edu/stat500/lesson/8/8.1

Party	Opinion
	favor	indifferent	opposed
Democrat	138	83	64
Bad	64	67	84

Gender and Condiment

Alpha: 0.1
Chi-score ~2.95
p-value ~0.229

source: https://online.stat.psu.edu/stat500/lesson/8/8.1

Gender	Condiment
	Ketchup	Mustard	Relish
Male	15	23	10
Female	25	19	8

estebanz01 / ruby-statistics