Comments by "T J" (@TJ-hs1qm) on "ThePrimeTime"
channel.
-
4
-
3
-
3
-
3
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
When I think about how to design a function I look at these factors
Can I cut down the dimensions of my function (as in dimension reduction)? Because reasoning about one input parameter is significantly easier than ten. I also take a look how my input values correlate. When they change, do they change independently or not. For example, when a person moves to different home, the street name, number, zip code and city tend to change simultaneously, but never their e-mail address. So maybe factor those out in something we call "Address Type" and "E-Mail Type" representing these uncorrelated multi dimensional concepts, so that I can pass in things that are highly correlated into my function (address and email), thereby reducing the number of input parameters and my cognitive load. After having figured out the degree of correlation between my input variables, I decide whether my function should really depend on two orthogonal variables, which could change for different reasons, meaning that my function is maybe doing too much. But all this always depends on the use case. If someone would move on to work for another company the work e-mail is probably also going to change with the address, which makes the e-mail correlate again with the address.
1
-
1
-
1
-
1
-
1