How do I store time-varying follower-following relationship?

Assume we have a small social media app, containing only 4 people, whos UID = 11, 22, 33, 44 respectively. These people would follow and unfollow each other over time, and say, I want to take note of their follower relationship on a weekly basis. In other words, I want to store temporal network relationship. Week 1, their follower network looks like: | primary serial | user own UID | followed UID | |---| --- | ---| | 1 | 11 | 22 | | 2 | 11 | 33 | | 3 | 22 | 33 | | 4 | 11 | 44 | Week 2, the change being that user 11 unfollowed 22, and user 44 followed 22: | primary serial | user own UID | followed UID | |---| --- | ---| | 1 | 11 | 33 | | 2 | 22 | 33 | | 3 | 11 | 44 | | 4 | 44 | 22 | Finally, now in Week 3, user 11 re-followed user 22: | primary serial | user own UID | followed UID | |---| --- | ---| | 1 | 11 | 22 | | 2 | 11 | 33 | | 3 | 22 | 33 | | 4 | 11 | 44 | | 5 | 44 | 22 | The 3 tables under each week's network are only for that specific week, but storing these week-tables creates redundancy, since some relations remain unchanged across time: | primary serial | user own UID | followed UID | |---| --- | ---| | ... | 11 | 33 | | ... | 22 | 33 | | ... | 11 | 44 | - So I wonder if there is a good way of keeping record of time-varying follower-following relationship across time, in 1 table rather than 3? - And given such a new (less-redundant) table, can I replicate the follower network structure of a given week directly (that is, I can get Week N network without needing to compute from Week 1 to Week N-1 networks)? - Should I add a column of week-specific labels? Should I add a boolean column to indicate whether or not the row's relation still exist? Any help would be highly appreciated! Thank you so much in advance!