Question

I have two tables that have the following structure:

Table 1: Show the number of tries a player has done.

Player_id, n_tries, date

Table 2: Show the number of hits a player has done.

Player_id, n_hits, date

At the end I want to have the % of hits (tries vs hits) a player has for today and the increment/decrement of hits between today and a week ago. And I would like the results to be ordered by the % of (today) hits descendent. So, for example, if we have a player with id = 1, and a player with id = 2, with this data:

Table 1:

1, 10, "2012-10-14"
2, 13, "2012-10-14"
1, 20, "2012-10-7"
2, 15, "2012-10-7"

Table 2:

1, 5, "2012-11-14"
2, 10, "2012-11-14"
1, 0, "2012-11-7"
2, 3, "2012-11-7"

I would need to finally show this:

Player_id, % hits (today), difference of hits (today vs a week ago)

2            77%       +233%
1            50%       +100% (actually is infinite -> (5 - 0)/0 )

I have the queries to get values from both the tables:

First table – Number of tries:

SELECT player_id, date, sum(n_tries) 
FROM player_tries 
WHERE (d_date = CURDATE() OR d_date = DATE_SUB(CURDATE(), INTERVAL 7 DAY))
GROUP BY date, player_id
ORDER BY player_id, date

I do a sum(n_tries) because a player can have multiple rows of tries for a particular day.

Second table – Number of hits:

SELECT player_id, date, sum(n_hits) 
FROM player_hits
WHERE (d_date = CURDATE() OR d_date = DATE_SUB(CURDATE(), INTERVAL 7 DAY))
GROUP BY date, player_id
ORDER BY player_id, date

So, my question is: If from the first query I get a list of result that look like this:

Player_id, date,         n_tries
1         2012-11-07        20
1         2012-11-14        10
2         2012-11-07        15
2         2012-11-14        13

And from the second table I get this:

 Player_id, date,         n_hits
    1         2012-11-07        0
    1         2012-11-14        5
    2         2012-11-07        3
    2         2012-11-14        10

What would be the best way to be able to mix those results and be able to have some kind of structure that is ordered by the % hits (today).?

Answer 1

Select hits from today and a week ago:

select h1.player_id, `h1.date`,
       sum(h1.hits) as hits_today,
       sum(h2.hits) as hits_last_week
from player_hits h1
left join player_hits h2 on h2.player_id = h1.player_id
                            and `h2.date` = date_sub(`h1.date`, interval 7 days)
group by h1.player_id, `h1.date`;

If you only want rows when there is a corresponding entry a week before, then replace the left join with an inner join.

The select for the tries table looks equivalent.

First select your results into arrays with player_id as keys.

$hits_array = array();
$result = mysqli_query(...);
while ($row = mysqli_fetch_assoc($result)) {
    $player = $row['player_id'];
    $hits_array[$player] = $row['hits'];
}

You do the same with the tries table.

Now you can join the two arrays on the player_id keys:

foreach ($hits_array as $player => $hits) {
    echo "$player " . ($hits / $tries[$key]) . "\n";
}