500is repeated twice, all generated values will use either score
500. The fields
game_idhave no repeated values, so everything generated will be random. Here's a possible result:
user_idhas samples between 1 and 72, so only values between 1 and 72 will be generated. Not only integers and floats can have ranges, date and datetime can have ranges, as well. So for example if you have a birthday field like in this sample data:
fromand a field
toand to is always bigger than from, than all the generated numbers will follow that rule. Currently
datetimevalues have detectedions for
=. An example:
fromis always smaller than
to, the result will look like this:
id1 and 5 have equal numbers and the row with
from. So no rule will be detected.
500is twice in the sample data and
250only once. This will be detected and the score 500 will have a higher chance of occuring in the generated data (twice as likely to be exact).
admin, so that field will be an optional one. Here's a generated dataset:
user_id. Now to have realistic values in those fields, you need to follow one simple rule, always define your field value range to the size of the table you're connecting to. Here's an example:
user_idrelationship field in the comments table should be synced to that range, so we add 1 and 50. As 250 comments will be generated, every user will have an average of five comments.