将新表用于性别或将其添加为人员表中的字段

Question

So basically which is better in terms of performance and space:

idPerson  Name  Gender   Age
1         John  Male    19
2         Anna  Female  22

or this way:

idPerson  Name  Age  idGender
1         John  19   1
2         Anna  22   2

idGender  Gender
1        Male
2        Female

I think the second one is better because by repeating Male/Female each time will consume more space but I think it may be slower cause of relation. I am wrong? There is better way?

Answer 1

which is better in terms of performance and space

Most decisions in Computer Science are usually a trade-off between time and space.

In your case, the first solution (storing a string up to 6 characters in the table) will ultimately use more space but results in that string being directly available to queries upon the table; whereas the second solution (storing an integer representation whose meaning can be looked up in a second table) will use less space but requires more work to discover the string (although it does enable one to restrict the possible values by defining a foreign key constraint ).

Another approach might be to use MySQL's ENUM type, which enables you to work with strings whilst effectively storing only an integer "under the hood". You may wish to consider Chris Komlenic's article 8 Reasons Why MySQL's ENUM Data Type Is Evil before deciding upon this course of action.

Answer 2

Assuming gender as a binary field, I would go with a third option: name the column MALE and the valid values will be either Y or N (or 1 / 0 )

Things to consider: other genders like transgender ;)

Answer 3

The best solution might be a combination of the both where you use a single character column inside the Person table to represent a Gender code that you map to in your application.

You can use ' M ' and ' F ' as codes for Male and Female, but now you're free to map any additional codes to whatever exotic genders you might think of.

Answer 4

I will go with first approach as it doesn't make much sense to split and keep gender information separate because its not going to change :). Had it been a scenario where gender name could change then i would have taken second approach.

EDIT: Adding this in response to below comment. A single table will be faster to query as you dont need any joins, so better performance. I dont think either approach will make much differnce in terms of space.