Python: two-way Anova

[bilhanbel]

I have a model in my task; one numerical 2 categorical variable. When I want to create a formula like you do here:
formula1 = 'numerical ~ C(cat1) + C(cat2)'

I can observe category one slightly less than 5% so I can reject null hypothesis however I see on another video they use one categorical variable to one numerical variable right?

so formula2 = 'numerical ~ cat1'
and I can observe that category one is 9%

what exactly means in difference here when we use these two dependent variables in formula1 and formula2
and which formula we use ?