Reputation: 23
I trying to get the last 6 months of the min and max of prices in my table and display them as a group by months. My query is not returning the corresponding rows values, such as the date time for when the max price was or min..
I want to select the min & max prices and the date time they both occurred and the rest of the data for that row... (the reason why i have concat for report_term, as i need to print this with the dataset when displaying results. e.g. February 2018 -> ...., January 2018 -> ...)
SELECT metal_price_id, CONCAT(MONTHNAME(metal_price_datetime), ' ', YEAR(metal_price_datetime)) AS report_term, max(metal_price) as highest_gold_price, metal_price_datetime FROM metal_prices_v2
WHERE metal_id = 1
AND DATEDIFF(NOW(), metal_price_datetime) BETWEEN 0 AND 180
GROUP BY report_term
ORDER BY metal_price_datetime DESC
I have made an example, extract from my DB: http://sqlfiddle.com/#!9/617bcb2/4/0
My desired result would be to see the min and max prices grouped by month, date of min, date of max.. and all in the last 6 months.
thanks
UPDATE. The below code works, but it returns back rows from beyond the 180 days specified. I have just checked, and it is because it joining by the price which may be duplicated a number of times during the years.... see: http://sqlfiddle.com/#!9/5f501b/1
Upvotes: 1
Views: 1916
Reputation: 48197
simplified version of what you should do so you can learn the working process.
You need calculate the min() max() of the periods you need. That is your first brick on this building.
you have tableA, you calculate min() lets call it R1
SELECT group_field, min() as min_value
FROM TableA
GROUP BY group_field
same for max() call it R2
SELECT group_field, max() as max_value
FROM TableA
GROUP BY group_field
Now you need to bring all the data from original fields so you join each result with your original table
We call those T1 and T2:
SELECT tableA.group_field, tableA.value, tableA.date
FROM tableA
JOIN ( ... .. ) as R1
ON tableA.group_field = R1.group_field
AND tableA.value = R1.min_value
SELECT tableA.group_field, tableA.value, tableA.date
FROM tableA
JOIN ( ... .. ) as R2
ON tableA.group_field = R2.group_field
AND tableA.value = R2.max_value
Now we join T1 and T2.
SELECT *
FROM ( .... ) as T1
JOIN ( .... ) as T2
ON t1.group_field = t2.group_field
So the idea is if you can do a brick, you do the next one. Then you also can add filters like last 6 months or something else you need.
In this case the group_field
is the CONCAT()
value
Upvotes: 0
Reputation: 133370
You could use twice inner join on the subselect for min and max
select a.metal_price_datetime
, t1.highest_gold_price
, t1.report_term
, t2.lowest_gold_price
,t2.metal_price_datetime
from metal_prices_v2 a
inner join (
SELECT CONCAT(MONTHNAME(metal_price_datetime), ' ', YEAR(metal_price_datetime)) AS report_term
, max(metal_price) as highest_gold_price
from metal_prices_v2
WHERE metal_id = 1
AND DATEDIFF(NOW(), metal_price_datetime) BETWEEN 0 AND 180
GROUP BY report_term
) t1 on t1.highest_gold_price = a.metal_price
inner join (
select a.metal_price_datetime
, t.lowest_gold_price
, t.report_term
from metal_prices_v2 a
inner join (
SELECT CONCAT(MONTHNAME(metal_price_datetime), ' ', YEAR(metal_price_datetime)) AS report_term
, min(metal_price) as lowest_gold_price
from metal_prices_v2
WHERE metal_id = 1
AND DATEDIFF(NOW(), metal_price_datetime) BETWEEN 0 AND 180
GROUP BY report_term
) t on t.lowest_gold_price = a.metal_price
) t2 on t2.report_term = t1.report_term
Upvotes: 3