Closed Many98 closed 3 years ago
1) I thought about doing just that but I think that if it is not so harming to the execution time, we should leave it be, because this dataframe is different in the purpose and the indexing. It stores data about a team and not a match. This way we can easily create a method or manipulate the frame differently in order to produce something like last N
matches of team XX
or for instance the matches of team XX
with team YY
. Is it a sufficient argument or do you still think we should get rid of it? (in summary, the self.matches
is meant to be a storage attribute from which we later will extract the feature attribute self.match_data
).
2) Ok, good point. I will rename it to self.LL_data
and self.SL_data
(unlike the self.LL_data
there will be a multiindex of 'Sea'
and 'team_ID
' (is it OK?))
c) I just made some changes that remove the self.last_match_data
completely and leave only the self.match_data
(this stores the date of the play so we can later easily access the last match result.
3) I encourage you to add any that you would feel usefull! ;)
Suggested changes are: