Let's say I have a table that reflects all of the individual purchases customers have made to date (see image below for the output i'm envisioning)
How would I write a query in MYSQL that returned these 2 columns, +:
A column that reflected the purchase date of that customer's purchase made directly prior (and in the case of no prior purchase, a null value)
A column that output a value of "1" for every difference in the two date columns that are greater than 70 days, a value of "0" for differences that are less than 70 days, and a null value for those that don't have a "prior purchase".
I have been working on this for days and I have only gotten it to work when I "GROUP BY" the customer ID's (using a self join that requires one date to be less than the other). I have no idea how i'd do it at the transaction level.
You can use a correlated subquery. Here is how you get the previous date:
select p.*,
(select p2.purchase_date
from purchases p2
where p2.customerid = p.customerid and
p2.purchase_date < p.purchase_date
order by p2.purchase_date desc
limit 1
) as prev_purchase_date
from purchases p;
You can use this as a subquery and then do the calculation for the final column using prev_purchase_date.
Related
I have users and orders tables with this structure (simplified for question):
USERS
userid
registered(date)
ORDERS
id
date (order placed date)
user_id
I need to get array of users (array of userid) who placed their 25th order during specified period (for example in May 2019), date of 25th order for each user, number of days to place 25th order (difference between registration date for user and date of 25th order placed).
For example if user registered in April 2018, then placed 20 orders in 2018, and then placed 21-30th orders in Jan-May 2019 - this user should be in this array, if he placed 25th (overall for his account) order in May 2019.
How I can do this with MySQL request?
Sample data and structure: http://www.sqlfiddle.com/#!9/998358 (for testing you can get 3rd order as ex., not 25th, to not add a lot of sample data records).
One request is not required - if this can't be done in one request, few is possible and allowed.
You can use a correlated subquery to get the count of orders placed before the current one by a user. If that's 24 the current order is the 25th. Then check if the date is in the desired range.
SELECT o1.user_id,
o1.date,
datediff(o1.date, u1.registered)
FROM orders o1
INNER JOIN users u1
ON u1.userid = o1.user_id
WHERE (SELECT count(*)
FROM orders o2
WHERE o2.user_id = o1.user_id
AND o2.date < o1.date
OR o2.date = o1.date
AND o2.id < o1.id) = 24
AND o1.date >= '2019-01-01'
AND o1.date < '2019-06-01';
The basic inefficient way of doing this would be to get the user_id for every row in ORDERS where the date is in your target range AND the count of rows in ORDERS with the same user_id and a lower date is exactly 24.
This can get very ugly, very quickly, though.
If you're calling this from code you control, can't you do it from the code?
If not, there should be a way to assign to each row an index describing its rank among orders for its specific user_id, and select from this all user_id from rows with an index of 25 and a correct date. This will give you a select from select from select, but it should be much faster. The difficulty here is to control the order of the rows, so here are the selects I envision:
Select all rows, order by user_id asc, date asc, union-ed to nothing from a table made of two vars you'll initialize at 0.
from this, select all while updating a var to know if a row's user_id is the same as the last, and adding a field that will report so (so for each user_id the first line in order will have a specific value like 0 while the other rows for the same user_id will have a 1)
from this, select all plus a field that equals itself plus one in case the first added field is 1, else 0
from this, select the user_id from the rows where the second added field is 25 and the date is in range.
The union thingy is only necessary if you need to do it all in one request (you have to initialize them in a lower select than the one they're used in).
Edit: Well if you need the date too you can just select it along with the user_id, but calculating the number of days in sql will be a pain. Just join the result table to the users table and get both the date of 25th order and their date of registration, you'll surely be able to do the difference in code.
I'll try building an actual request, however if you want to truly understand what you need to make this you gotta read up on mysql variables, unions, and conditional statements.
"Looks too complicated. I am sure that this can be done with current DB structure and 1-2 requests." Well, yeah. Use the COUNT request, it will be easy, and slow as hell.
For the complex answer, see http://www.sqlfiddle.com/#!9/998358/21
Since you can use multiple requests, you can just initialize the vars first.
It isn't actually THAT complicated, you just have to understand how to concretely express what you mean by "an user's 25th command" to a SQL engine.
See http://www.sqlfiddle.com/#!9/998358/24 for the difference in days, turns out there's a method for that.
Edit 5: seems you're going with the COUNT method. I'll pray your DB is small.
Edit 6: For posterity:
The count method will take years on very large databases. Since OP didn't come back, I'm assuming his is small enough to overlook query speed. If that's not your case and let's say it's 10 years from now and the sqlfiddle links are dead; here's the two-queries solution:
SET #PREV_USR:=0;
SELECT user_id, date_ FROM (
SELECT user_id, date_, SAME_USR AS IGNORE_SMUSR,
#RANK_USR:=(CASE SAME_USR WHEN 0 THEN 1 ELSE #RANK_USR+1 END) AS RANK FROM (
SELECT orders.*, CASE WHEN #PREV_USR = user_id THEN 1 ELSE 0 END AS SAME_USR,
#PREV_USR:=user_id AS IGNORE_USR FROM
orders
ORDER BY user_id ASC, date_ ASC, id ASC
) AS DERIVED_1
) AS DERIVED_2
WHERE RANK = 25 AND YEAR(date_) = 2019 AND MONTH(date_) = 4 ;
Just change RANK = ? and the conditions to fit your needs. If you want to fully understand it, start by the innermost SELECT then work your way high; this version fuses the points 1 & 2 of my explanation.
Now sometimes you will have to use an API or something and it wont let you keep variable values in memory unless you commit it or some other restriction, and you'll need to do it in one query. To do that, you put the initialization one step lower and make it so it does not affect the higher statements. IMO the best way to do this is in a UNION with a fake table where the only row is excluded. You'll avoid the hassle of a JOIN and it's just better overall.
SELECT user_id, date_ FROM (
SELECT user_id, date_, SAME_USR AS IGNORE_SMUSR,
#RANK_USR:=(CASE SAME_USR WHEN 0 THEN 1 ELSE #RANK_USR+1 END) AS RANK FROM (
SELECT DERIVED_4.*, CASE WHEN #PREV_USR = user_id THEN 1 ELSE 0 END AS SAME_USR,
#PREV_USR:=user_id AS IGNORE_USR FROM
(SELECT * FROM orders
UNION
SELECT * FROM (
SELECT (#PREV_USR:=0) AS INIT_PREV_USR, 0 AS COL_2, 0 AS COL_3
) AS DERIVED_3
WHERE INIT_PREV_USR <> 0
) AS DERIVED_4
ORDER BY user_id ASC, date_ ASC, id ASC
) AS DERIVED_1
) AS DERIVED_2
WHERE RANK = 25 AND YEAR(date_) = 2019 AND MONTH(date_) = 4 ;
With that method, the thing to watch for is the amount and the type of columns in your basic table. Here orders' first field is an int, so I put INIT_PREV_USR in first then there are two more fields so I just add two zeroes with names and call it a day. Most types work, since the union doesn't actually do anything, but I wouldn't try this when your first field is a blob (worst comes to worst you can use a JOIN).
You'll note this is derived from a method of pagination in mysql. If you want to apply this to other engines, just check out their best pagination calls and you should be able to work thinks out.
I currently have an employee logging sql table that has 3 columns
fromState: String,
toState: String,
timestamp: DateTime
fromState is either In or Out. In means employee came in and Out means employee went out. Each row can only transition from In to Out or Out to In.
I'd like to generate a temporary table in sql to keep track during a given hour (hour by hour), how many employees are there in the company. Aka, resulting table has columns HourBucket, NumEmployees.
In non-SQL code I can do this by initializing the numEmployees as 0 and go through the table row by row (sorted by timestamp) and add (employee came in) or subtract (went out) to numEmployees (bucketed by timestamp hour).
I'm clueless as how to do this in SQL. Any clues?
Use a COUNT ... GROUP BY query. Can't see what you're using toState from your description though! Also, assuming you have an employeeID field.
E.g.
SELECT fromState AS 'Status', COUNT(*) AS 'Number'
FROM StaffinBuildingTable
INNER JOIN (SELECT employeeID AS 'empID', MAX(timestamp) AS 'latest' FROM StaffinBuildingTable GROUP BY employeeID) AS LastEntry ON StaffinBuildingTable.employeeID = LastEntry.empID
GROUP BY fromState
The LastEntry subquery will produce a list of employeeIDs limited to the last timestamp for each employee.
The INNER JOIN will limit the main table to just the employeeIDs that match both sides.
The outer GROUP BY produces the count.
SELECT HOUR(SBT.timestamp) AS 'Hour', SBT.fromState AS 'Status', COUNT(*) AS 'Number'
FROM StaffinBuildingTable AS SBT
INNER JOIN (
SELECT SBIJ.employeeID AS 'empID', MAX(timestamp) AS 'latest'
FROM StaffinBuildingTable AS SBIJ
WHERE DATE(SBIJ.timestamp) = CURDATE()
GROUP BY SBIJ.employeeID) AS LastEntry ON SBT.employeeID = LastEntry.empID
GROUP BY SBT.fromState, HOUR(SBT.timestamp)
Replace CURDATE() with whatever date you are interested in.
Note this is non-optimal as it calculates the HOUR twice - once for the data and once for the group.
Again you are using the INNER JOIN to limit the number of returned row, this time to the last timestamp on a given day.
To me your description of the FromState and ToState seem the wrong way round, I'd expect to doing this based on the ToState. But assuming I'm wrong on that the following should point you in the right direction:
First, I create a "Numbers" table containing 24 rows one for each hour of the day:
create table tblHours
(Number int);
insert into tblHours values
(0),(1),(2),(3),(4),(5),(6),(7),
(8),(9),(10),(11),(12),(13),(14),(15),
(16),(17),(18),(19),(20),(21),(22),(23);
Then for each date in your employee logging table, I create a row in another new table to contain your counts:
create table tblDailyHours
(
HourBucket datetime,
NumEmployees int
);
insert into tblDailyHours (HourBucket, NumEmployees)
select distinct
date_add(date(t.timeStamp), interval h.Number HOUR) as HourBucket,
0 as NumEmployees
from
tblEmployeeLogging t
CROSS JOIN tblHours h;
Then I update this table to contain all the relevant counts:
update tblDailyHours h
join
(select
h2.HourBucket,
sum(case when el.fromState = 'In' then 1 else -1 end) as cnt
from
tblDailyHours h2
join tblEmployeeLogging el on
h2.HourBucket >= el.timeStamp
group by h2.HourBucket
) cnt ON
h.HourBucket = cnt.HourBucket
set NumEmployees = cnt.cnt;
You can now retrieve the counts with
select *
from tblDailyHours
order by HourBucket;
The counts give the number on site at each of the times displayed, if you want during the hour in question, we'd need to tweak this a little.
There is a working version of this code (using not very realistic data in the logging table) here: rextester.com/DYOR23344
Original Answer (Based on a single over all count)
If you're happy to search over all rows, and want the current "head count" you can use this:
select
sum(case when t.FromState = 'In' then 1 else -1) as Heads
from
MyTable t
But if you know that there will always be no-one there at midnight, you can add a where clause to prevent it looking at more rows than it needs to:
where
date(t.timestamp) = curdate()
Again, on the assumption that the head count reaches zero at midnight, you can generalise that method to get a headcount at any time as follows:
where
date(t.timestamp) = "CENSUS DATE" AND
t.timestamp <= "CENSUS DATETIME"
Obviously you'd need to replace my quoted strings with code which returned the date and datetime of interest. If the headcount doesn't return to zero at midnight, you can achieve the same by removing the first line of the where clause.
I have several years worth of data in a table (inquiries). Every entry has a contact_time field that is the timestamp of their email contact. I'm trying to build monthly or weekly summary data for plotting on a multi-series graph. To that end, I need to see the month or week number in the first column with the respective data from 2014 in the second column, and from 2015 in the third column, etc.
SELECT MONTH(inquiries.contact_time) AS "Date",
(SELECT
COUNT(inquiries.id) AS "Inquiries"
FROM inquiries
WHERE YEAR(inquiries.contact_time) = "2014"
) AS "2014",
(SELECT COUNT(inquiries.id) AS "Inquiries"
FROM inquiries
WHERE YEAR(inquiries.contact_time) = "2015"
) AS "2015"
FROM inquiries
GROUP BY MONTH(inquiries.contact_time)
All I'm seeing is the total count for each year in all of the rows. Any help is appreciated.
Use conditional aggregation:
SELECT MONTH(i.contact_time),
SUM(YEAR(i.contact_time) = 2014) as cnt_2014,
SUM(YEAR(i.contact_time) = 2015) as cnt_2015
FROM inquiries i
WHERE YEAR(i.contact_time) >= 2014
GROUP BY MONTH(i.contact_time);
If you have an index on contact_time, then use the condition where i.contact_time >= '2014-01-01', so it can take advantage of the index.
You're seeing the total count for the year because your subqueries are not related to the month for the outer query's grouping.
I would write the query this way:
SELECT MONTH(contact_time) AS `Date`,
SUM(YEAR(contact_time)=2014) AS `2014`,
SUM(YEAR(contact_time)=2015) AS `2015`
FROM inquiries
GROUP BY MONTH(contact_time)
Explanation: the COUNT() of a specific set of rows is the same as the SUM() of 1's for those rows. And MySQL boolean expressions return the integer 1 for true.
The question I am working on is as follows:
What is the difference in the amount received for each month of 2004 compared to 2003?
This is what I have so far,
SELECT #2003 = (SELECT sum(amount) FROM Payments, Orders
WHERE YEAR(orderDate) = 2003
AND Payments.customerNumber = Orders.customerNumber
GROUP BY MONTH(orderDate));
SELECT #2004 = (SELECT sum(amount) FROM Payments, Orders
WHERE YEAR(orderDate) = 2004
AND Payments.customerNumber = Orders.customerNumber
GROUP BY MONTH(orderDate));
SELECT MONTH(orderDate), (#2004 - #2003) AS Diff
FROM Payments, Orders
WHERE Orders.customerNumber = Payments.customerNumber
Group By MONTH(orderDate);
In the output I am getting the months but for Diff I am getting NULL please help. Thanks
I cannot test this because I don't have your tables, but try something like this:
SELECT a.orderMonth, (a.orderTotal - b.orderTotal ) AS Diff
FROM
(SELECT MONTH(orderDate) as orderMonth,sum(amount) as orderTotal
FROM Payments, Orders
WHERE YEAR(orderDate) = 2004
AND Payments.customerNumber = Orders.customerNumber
GROUP BY MONTH(orderDate)) as a,
(SELECT MONTH(orderDate) as orderMonth,sum(amount) as orderTotal FROM Payments, Orders
WHERE YEAR(orderDate) = 2003
AND Payments.customerNumber = Orders.customerNumber
GROUP BY MONTH(orderDate)) as b
WHERE a.orderMonth=b.orderMonth
Q: How do I subtract two declared variables in MySQL.
A: You'd first have to DECLARE them. In the context of a MySQL stored program. But those variable names wouldn't begin with an at sign character. Variable names that start with an at sign # character are user-defined variables. And there is no DECLARE statement for them, we can't declare them to be a particular type.
To subtract them within a SQL statement
SELECT #foo - #bar AS diff
Note that MySQL user-defined variables are scalar values.
Assignment of a value to a user-defined variable in a SELECT statement is done with the Pascal style assignment operator :=. In an expression in a SELECT statement, the equals sign is an equality comparison operator.
As a simple example of how to assign a value in a SQL SELECT statement
SELECT #foo := '123.45' ;
In the OP queries, there's no assignment being done. The equals sign is a comparison, of the scalar value to the return from a subquery. Are those first statements actually running without throwing an error?
User-defined variables are probably not necessary to solve this problem.
You want to return how many rows? Sounds like you want one for each month. We'll assume that by "year" we're referring to a calendar year, as in January through December. (We might want to check that assumption. Just so we don't find out way too late, that what was meant was the "fiscal year", running from July through June, or something.)
How can we get a list of months? Looks like you've got a start. We can use a GROUP BY or a DISTINCT.
The question was... "What is the difference in the amount received ... "
So, we want amount received. Would that be the amount of payments we received? Or the amount of orders that we received? (Are we taking orders and receiving payments? Or are we placing orders and making payments?)
When I think of "amount received", I'm thinking in terms of income.
Given the only two tables that we see, I'm thinking we're filling orders and receiving payments. (I probably want to check that, so when I'm done, I'm not told... "oh, we meant the number of orders we received" and/or "the payments table is the payments we made, the 'amount we received' is in some other table"
We're going to assume that there's a column that identifies the "date" that a payment was received, and that the datatype of that column is DATE (or DATETIME or TIMESTAMP), some type that we can reliably determine what "month" a payment was received in.
To get a list of months that we received payments in, in 2003...
SELECT MONTH(p.payment_received_date)
FROM payment_received p
WHERE p.payment_received_date >= '2003-01-01'
AND p.payment_received_date < '2004-01-01'
GROUP BY MONTH(p.payment_received_date)
ORDER BY MONTH(p.payment_received_date)
That should get us twelve rows. Unless we didn't receive any payments in a given month. Then we might only get 11 rows. Or 10. Or, if we didn't receive any payments in all of 2003, we won't get any rows back.
For performance, we want to have our predicates (conditions in the WHERE clause0 reference bare columns. With an appropriate index available, MySQL will make effective use of an index range scan operation. If we wrap the columns in a function, e.g.
WHERE YEAR(p.payment_received_date) = 2003
With that, we will be forcing MySQL to evaluate that function on every flipping row in the table, and then compare the return from the function to the literal. We prefer not do do that, and reference bare columns in predicates (conditions in the WHERE clause).
We could repeat the same query to get the payments received in 2004. All we need to do is change the date literals.
Or, we could get all the rows in 2003 and 2004 all together, and collapse that into a list of distinct months.
We can use conditional aggregation. Since we're using calendar years, I'll use the YEAR() shortcut (rather than a range check). Here, we're not as concerned with using a bare column inside the expression.
SELECT MONTH(p.payment_received_date) AS `mm`
, MAX(MONTHNAME(p.payment_received_date)) AS `month`
, SUM(IF(YEAR(p.payment_received_date)=2004,p.payment_amount,0)) AS `2004_month_total`
, SUM(IF(YEAR(p.payment_received_date)=2003,p.payment_amount,0)) AS `2003_month_total`
, SUM(IF(YEAR(p.payment_received_date)=2004,p.payment_amount,0))
- SUM(IF(YEAR(p.payment_received_date)=2003,p.payment_amount,0)) AS `2004_2003_diff`
FROM payment_received p
WHERE p.payment_received_date >= '2003-01-01'
AND p.payment_received_date < '2005-01-01'
GROUP
BY MONTH(p.payment_received_date)
ORDER
BY MONTH(p.payment_received_date)
If this is a homework problem, I strongly recommend you work on this problem yourself. There are other query patterns that will return an equivalent result.
I think this is the problem:
In #2003 and #2004, you select only the sum. And even if you group by the month you still select one column i.e. each row does not say what month it is select for. So when you try to subtract SQL asks which row in #2003 should be subtracted from #2004.
So I think the solution is to select the month with the sum and do the subtract later based on the month.
I'm looking to make some bar graphs to count item sales by day, month, and year. The problem that I'm encountering is that my simple MySQL queries only return counts where there are values to count. It doesn't magically fill in dates where dates don't exist and item sales=0. This is causing me problems when trying to populate a table, for example, because all weeks in a given year aren't represented, only the weeks where items were sold are represented.
My tables and fields are as follows:
items table: account_id and item_id
// table keeping track of owners' items
items_purchased table: purchaser_account_id, item_id, purchase_date
// table keeping track of purchases by other users
calendar table: datefield
//table with all the dates incremented every day for many years
here's the 1st query I was referring to above:
SELECT COUNT(*) as item_sales, DATE(purchase_date) as date
FROM items_purchased join items on items_purchased.item_id=items.item_id
where items.account_id=125
GROUP BY DATE(purchase_date)
I've read that I should join a calendar table with the tables where the counting takes place. I've done that but now I can't get the first query to play nice this 2nd query because the join in the first query eliminates dates from the query result where item sales are 0.
here's the 2nd query which needs to be merged with the 1st query somehow to produce the results i'm looking for:
SELECT calendar.datefield AS date, IFNULL(SUM(purchaseyesno),0) AS item_sales
FROM items_purchased join items on items_purchased.item_id=items.item_id
RIGHT JOIN calendar ON (DATE(items_purchased.purchase_date) = calendar.datefield)
WHERE (calendar.datefield BETWEEN (SELECT MIN(DATE(purchase_date))
FROM items_purchased) AND (SELECT MAX(DATE(purchase_date)) FROM items_purchased))
GROUP BY date
// this lists the sales/day
// to make it per week, change the group by to this: GROUP BY week(date)
The failure of this 2nd query is that it doesn't count item_sales by account_id (the person trying to sell the item to the purchaser_account_id users). The 1st query does but it doesn't have all dates where the item sales=0. So yeah, frustrating.
Here's how I'd like the resulting data to look (NOTE: these are what account_id=125 has sold, other people many have different numbers during this time frame):
2012-01-01 1
2012-01-08 1
2012-01-15 0
2012-01-22 2
2012-01-29 0
Here's what the 1st query current looks like:
2012-01-01 1
2012-01-08 1
2012-01-22 2
If someone could provide some advice on this I would be hugely grateful.
I'm not quite sure about the problem you're getting as I don't know the actual tables and data they contain that generates those results (that would help a lot!). However, let's try something. Use this condition:
where (items.account_id = 125 or items.account_id is null) and (other-conditions)
Your first query is perfectly acceptable. The fact is you don't have data in the mysql table and therefore it can't group any data together. This is fine. You can account for this in your code so that if the date does not exist, then obviously there's no data to graph. You can better account for this by ordering the date value so you can loop through it accordingly and look for missed days.
Also, to avoid doing the DATE() function, you can change the GROUP BY to GROUP BY date (because you have in your fields selected DATE(pruchase_date) as date)