I stumbled across this website and instantly fell in love. Let me be completely honest, I have little to NO knowledge of Access.. I told my manager this and he still insists that I "can figure it out" which is highly doubted. So here I am asking for help.. On to the question:
Where are the SQL code gurus? haha
I have 2 tables, "Found" & "Missing", both showing inventory adjustments for our building within the company. (Amazon)
I believe I have the process figured out but have no idea how it looks within Access..
Step 1: Group by ASIN (basically the numerical version of a barcode)
Step 2: Determine the +/- for the grouped ASINs in both lists
Step 3: Use TOP function to find the largest negative adjustments
There is a total of 3000+ records in both spreadsheets, but hopefully if I can figure out the process then the input/output wouldn't matter.
I thought maybe I needed a unique identifier? Bin(location) + ASIN(barcode) + Quantity
As you can see.. I have been thinking, organizing, and praying someone can help!
Here is a dummy example of the "Found" spreadsheet, the "Missing" spreadsheet is the exact same format with the only difference being a "M" instead of a "F" under "Reason Code"
Hopefully this is enough information, I know its a cluster.... thanks guys!
Date FC Application Name IOG ID IOG Name Container Id GL Product Group ASIN Processed By Reason Code Quantity Item Cost
1/5/2014 RIC1 FCICQACountService 1234 Doll Inc. P-1-A101xxx Toy B000000001 unknown1 F -1 12.34
1/5/2014 RIC1 FCICQACountService 1334 Amazon P-1-A101xxx Drugstore B000000002 unknown2 F -1 10.36
1/5/2014 RIC1 FCICQACountService 1432 Amazon P-1-A102xxx Office Product B000000003 unknown3 F -13 50.50
1/5/2014 RIC1 FCICQACountService 1442 Amazon P-1-A102xxx Office Product B000000004 unknown4 F -2 223.62
1/5/2014 RIC1 FCICQACountService 1337 Hope Inc. P-1-A102xxx Office Product B000000005 unknown5 F -1 100.99
I take it that by "spreadsheet", you actually mean "table". Might be a good idea to find a good primer on SQL and relational databases in general.
You've got a pretty good start, though. You've identified what you want. Note that in SQL, this is what you usually do; you think more about the result you want than you do the process of getting it. Each of your points suggests a keyword or function that will go into your query:
1) "Group by ASIN (basically the numerical version of a barcode)": You probably want to use the GROUP BY keyword.
2) "Determine the +/- for the grouped ASINs in both lists": Sounds like you want to SUM up a column here.
3) "Use TOP function to find the largest negative adjustments": Obviously, you already know you want TOP. The piece you're missing, though, is that the "largest negative" part suggests you want to use ORDER BY, and you want the smallest (largest magnitude negative) first. That will make sure that the right row is on top when it takes the top one.
So putting all that together, the only thing you need to figure out is the syntax. Your end query probably looks something like this:
SELECT TOP 1 ASIN, SUM(Quantity) AS TotalQuantity FROM Found GROUP BY ASIN ORDER BY SUM(Quantity);
This will calculate the sum of Quantity for each group of rows that has the same ASIN, and the result will be a set of rows that contain the ASIN and the total Quantity for that ASIN. Then it sorts the rows using the total quantity, with the smallest (most negative) row on top. The TOP then cuts off all the other rows. You could optionally leave out the TOP 1 if you want to see all the rows.
By the way, this SUM function is a little special. It's what we call an aggregate function. That's because it does something with a bunch of values across many rows. Not all functions are like that in SQL, but this one is.
If this isn't exactly what you're looking for, I hope it's enough to get you off the ground. Good luck.
Related
I try to fix an SQL i did Earlier and have a problem while doing that.
I have an Huge SQL witch Calculate SUM Fields for a Report.
ATM The SQL Counts Every Vehicle, even when its a Electro Car.
CASE
WHEN AVG(mov.TOTALCONSUMPTION) >0
AND AVG((moving.TOTALTIME)) >0
THEN ROUND(AVG(mov.TOTALCONSUMPTION) / (AVG(mov.TOTALTIME)/3600), 4)
ELSE 0
END ConsumptionPerHour,
I need the Condition Where vehicle.energysource = 0 in my sql,
I try to build it in but allways get the err. "not a GROUP BY expression"
CASE
WHEN AVG(mov.TOTALCONSUMPTION) >0
AND AVG((moving.TOTALTIME)) >0
AND vehicle.ENERGYSOURCE= 0
THEN ROUND(AVG(mov.TOTALCONSUMPTION) / (AVG(mov.TOTALTIME)/3600), 4)
ELSE 0
END ConsumptionPerHour,
I know what the Error Mean but i have no Idear how to fix this.
Anyone have a Idear how i can solve this Problem?
(Btw its Oracle sql developer if that is helpfull)
Edit:
My Exact problem is that i cant put it in the Where part becouse my SQL have to Calculate some things with Electro Cars too like The driven Killometers Ect.
I have the Fuel Problem becouse i want to calculate the avg.
Lets say Car 1 is a Electro Car and Car to a Diesel Car.
The 1 one have a Consumption of 0(becouse Electric car) and the 2. one of 20.
My SQL now Calculate Sum of Consumption (Car 1+Car2 = 20) / amount of Cars:
Thats make a Avg Consumption of 10.
But i only need the Avg Consumption of all cars who are not electric cars.
Maybe a noob Question but im a Trainee and alone in the Office, so sorry if thats an Easy or bad explained question.
Okay i found a Solution, i simple created a Sub-Select and filter here if
vd.fuel = 0
Anyways, Thanks for Help.
I have a database in which there are 13 different products, sold in 6 different countries.
Prices increase once a year.
Prices need to be calculated using a linear interpolation method. I have 21 different price and quantity increments for each product for each country for each year.
The user needs to be able to see how much an order would cost for any given value (as you would expect).
What the database needs to do (in English!) is to:
If there is a matching quantity from TblOrderDetail in the TblPrices,
use the price for the current product, country and year
if there isn't a matching quantity but the quantity required is greater than 1000 for one product (GT) and greater than 100 for every other product:
Find the highest quantity for the product, country and year (so, 1000 or 100, depending on the product), and calculate a pro-rated price. eg. If someone wanted 1500 of product GT for the UK for 2015, we'd look at the price for 1000 GT in the UK for 2015 and multiply it by 1.5. If 1800 were required, we'd multiply it by 1.8. I haven't been able to get this working yet as I'm looking at it alongside the formula for the next possibility...
If there isn't a matching quantity and the quantity required is less than 1000 for the product GT but 100 for the other products (this is the norm)...
Find the quantity and price for the increment directly below the quantity required by the user for the required product, country and year (let's call these quantitybelow and pricebelow)
Find the quantity and price for the increment directly above the quantity required by the user for the required product, country and year (let's call these quantityabove and priceabove)
Calculate the price for the required number of products for an account holder in a particular country for a given year using this formula.
ActualPrice: PriceBelow + ((PriceAbove - PriceBelow) * (The quantity required in the order detail - QuantityBelow) / (QuantityAbove - QuantityBelow))
I have spent days on this and have sought advice about this before but I am still getting very stuck.
The tables I've been working with to try and make this work are as follows:
TblAccount (primary key is AccountID, it also has a Country field which joins to the TblCountry.Code (primary key)
TblOrders (primary key is Order ID) which joins to TblAccount via the AccountID field; TblOrderDetail via the OrderID. This table also holds the OrderDate and Recipient ID which links to a person in TblContact - I don't need that here but will need it later to generate an invoice
TblOrderDetail (primary key is DetailID) which joins to TblOrders via OrderID field; TblProducts via ProductID field, and holds the Quantity required as well as the product
TblProducts (primary key is ProductCode) which as well as joining to TblOrderDetail, also joins to TblPrice via the Product field
TblPrices links to the TblProducts (as you have just read). I've also created an Alias for the TblCountry (CountryAliasForProductCode) so I can link it to the TblPrices to show the country link. I'm not sure if I needed to do this - it doesn't work if I do or I don't do it, so I seek guidance again here.
This is the code I've been trying to use (and failing) to get my price and quantity steps above and I hope to replicate it, making a couple of tweaks to get the steps below:
SELECT MIN(TblPrices.stepquantity) AS QuantityAbove, MIN(TblPrices.StepPrice) AS PriceAbove, TblOrders.OrderID, TblOrders.OldOrderID, TblOrders.AccountID, TblOrders.OrderDate, TblOrders.RecipientID, TblOrders.OrderStatus, TblOrderDetail.DetailID, TblOrderDetail.Product, TblOrderDetail.Quantity
FROM (TblCountry INNER JOIN ((TblAccount INNER JOIN TblOrders ON TblAccount.AccountID = TblOrders.AccountID) INNER JOIN (TblOrderDetail INNER JOIN TblProducts ON TblOrderDetail.Product = TblProducts.ProductCode) ON TblOrders.OrderID = TblOrderDetail.OrderID) ON TblCountry.Code = TblAccount.Country) INNER JOIN (TblCountry AS CountryAliasForProduct INNER JOIN TblPrices ON CountryAliasForProduct.Code = TblPrices.CountryCode) ON TblProducts.ProductCode = TblPrices.Product
WHERE (StepQuantity >= TblOrderDetails.Quantity)
AND (TblPrices.CountryCode = TblAccount.Country)
AND (TblOrderDetail.Product = TblPrices.Product)
AND (DATEPART('yyyy', TblPrices.DateEffective) = DATEPART('yyyy', TblOrders.OrderDate));
I've also tried...
I've even tried going back to basics and trying again to generate the steps below in 1 query, then try the steps above in another and finally, create the final calculation in another query.
This is what I have been trying to get my prices and quantities below:
SELECT Max(StepQuantity) AS quantity_below, Max(StepPrice) AS price_below, TblOrderDetails.Quantity, TblAccounts.Country
FROM
(TblProducts INNER JOIN TblPrices ON TblProducts.ProductCode = TblPrices.Product)
(TblOrderDetail INNER JOIN TblProducts ON TblOrderDetail.Product = TblProducts.ProductCode)
(TblOrders INNER JOIN TblOrderDetail ON TblOrders.OrderID = TblOrderDetail.OrderID)
(TblAccount INNER JOIN TblOrders ON TblAccount.AccountID = TblOrders.AccountID),
WHERE (((TblPrices.StepQuantity)<=(TblOrderDetail.Quantity)) AND ((TblPrices.CountryCode)=([TblAccounts].[country])) AND ((TblPrices.Product)=([TblOrderDetail].[product])) AND ((DatePart('yyyy',[TblPrices].[DateApplicable]))=(DatePart('yyyy',[TblOrders].[OrderDate]))));
You may be able to see glaring errors in this but I'm afraid I can't. I've tried re-jigging it and I'm getting nowhere.
I need to be able to tie the information in to the OrderDetail records as the price generated will need to be added to a financial transactions table as a debit amount and will show as an amount owing on statements.
I'm really not very good at SQL. I've read and worked though several self-study books and I have asked part of this question before; but I really am struggling with it. If anyone has any ideas on how to proceed, or even where I've gone wrong with my code, I'd be delighted, even if you tell me I shouldn't be using SQL. For the record, I originally posted this question on a different forum under Visual Basic. Responses from that forum brought me to SQL - however, anything that works would be good!
I've even tried, using Excel, concatenating the Year&Product&Country&Quantity to get a unique product code, interpolating the prices for every quantity between 1 and 1000 for each product, country and year and bringing them into a TblProductsAndPrices table. In Access, I created a query to concatenate the Year(of order date from tblOrders)&Product(of tblorderdetails)&Country(of tblAccount) in order to get the required product code for the order. Another query would find a price for me. However, any product code that doesn't appear on the list (such as where a quantity isn't listed in the tblProductsAndPrices as it is larger than the highest price increment) doesn't have a price.
If there was a workable solution to what I've just described that would generate a price for everything, then I'd be so pleased.
I'd really like to be able to generate an order for any quantity of any product for any account based in any country on any date and retrieve a price which will be used to "debit" a financial account in the database, who in a transaction history for an account and appear on statements. I'd also like to be able to do an ad-hoc price check on the spot.
Thank you very much for taking the time to read this. I really appreciate it. If you could offer any help or words of encouragement, I'd be very grateful.
Many thanks
Karen
Maybe no one thinks on an easy solution to the problem, since not all minds work in database thinking.
Easy solution: Create one view that gives all calculated values, not only the final one you need, each one as a column. Then you can use such view in a relation view and use on some rows one of the values and on other rows other values, etc.
How to think is simple, think in reverse order, instead of thinking "if that then I need to calculate such else I need this other", think as "I need "such" and I need "this other", both are columns of an intermediate view, then think on top level "if" that would be another view, such view will select the correct value ignoring the rest.
Never ever try to solve all in one step, that can be a really big headache.
Pros: You can isolate calculated values (needed or not), sql is much more easy to write and maintain.
Cons: Resources use is bigger than minimal, but most of times that extra calculated values does not represent a really big impact.
In terms of tutorial out there: Instead of a Top-Down method, use a Down-Top method.
Sometimes it is better (with your example) to calculate all three values (you write sentences on bold) ignoring the if part, and have all three possible values for your order and after that discard the ones not wanted, than trying to only calculate one.
Trying to calculate only one is thinking as a procedural programming, when working with databases most times one must get rid of such thinking and think as reverse, first do the most internal part of such procedural programming to have all data collected, then do the external selection of the procedural programing.
Note: If one of the values can not be calculated, just generate a Null.
I know it is hard to think on First in, last out (Down-Top) model, but it is great for things as the one you want.
Step1 (on specific view, or a join from one view per calculation):
Calculate column 1 as price for the current product, country and
year
Calculate column 2 as calculate a pro-rated price as if 1000
Calculate column 3 as calculate a pro-rated price as if 100
Calculate column 4 as etc
Calculate column N as etc
Step 2 (Another view, the one you want):
Calculate the if part, so you can choose adequate column from previous view (you can use immediately if or a calculated auxiliary field).
Hope you can follow theese way of thinking, I have solved a lot of things like that one (and more complex) thinking in that way, but it is not easy to think as that, needs an extra effort.
My background in SQL is limited, but my Googling isn't, I feel that I might just be missing the vocabulary to ask this question properly so hopefully beyond an answer to my question I can get the vocabulary I need to research this issue further.
I have a parts table - PARTS
I have a Purchase Order table - PO
and I have a PO Line Item table - PO_LINEITEM
The question I'm attempting to answer is given a particular part I want to get the latest purchase order and then look at the price we paid.
The PO table holds the date (PO_DATE) of when the Purchase Order was filled and the PO_LINEITEM table holds the information regarding the particular line item such as part primary key (PART_PRIMARYKEY) and price (PART_PRICE). The PARTS table isn't as important as the rest except that I need to return a PART primary key so that I can base the resulting view off the PARTS table
I've been through sub-queries and scalable sub-queries and the like, but I can't seem to find the right combination.
I started from a base of:
SELECT a.PO_DATE, b.PART_PRIMARYKEY, b.PART_PRICE
FROM PO a, PO_LINEITEM b
WHERE a.PO_PRIMARYKEY = b.PO_PRIMARYKEY
As you would expect this returns a list of every instance of an object in a PO with it's price and the date the Purchase Order was filled. The closest I have come to crack this is by using the MAX function on the date such as:
SELECT MAX(a.PO_DATE) AS DATE, b.PART_PRIMARYKEY, b.PART_PRICE
FROM PO a, PO_LINEITEM b
WHERE a.PO_PRIMARYKEY = b.PO_PRIMARYKEY
GROUP BY b.PART_PRIMARYKEY, b.PART_PRICE
This returns a Max date for each price a we paid for a particular part, so:
PART 1234, £12.95, 12/08/2012
PART 1234, £13.00, 14/08/2012
PART 1234. £11.15, 17/08/2012
PART 2345, £5.25, 12/08/2012
PART 2345, £5.65, 13/08/2012
etc.
What I need is:
PART 1234, £11.15, 17/08/2012
PART 2345. £5.65, 13/08/2012
If I could just group by the PART_PRIMARYKEY that would be excellent, but I get an ORA-00979 not a GROUP BY expression when I try.
Like I said I feel that my lack of vocabulary around this issue is impeding me finding an answer, so if anyone could point me in the right direction I'd be grateful
So hopefully I'm not asking a question that is asked every other day, but haven't found because I didn't use the magical combination of words to find.
Thank you for any help you can offer.
Look up Analytic Functions. They were introduced in 8i though I'm not sure how advanced they were at the time compared to how very good they can be in 11. A few links that I've used to understand them:
http://www.oracle-base.com/articles/misc/analytic-functions.php
http://www.orafaq.com/node/55
Though, a sub-query such as this might suffice (I may have your column naming mixed up):
select A.PART_PRIMARYKEY, B.PART_PRICE, A.PO_DATE
from PO A, PO_LINEITEM B
where A.PO_PRIMARYKEY = B.PO_PRIMARYKEY
and (A.PART_PRIMARYKEY, A.PO_DATE) in
( select A.PART_PRIMARYKEY, max(A.PO_DATE)
from PO A, PO_LINEITEM B
where A.PO_PRIMARYKEY = B.PO_PRIMARYKEY
group by A.PART_PRIMARYKEY);
I've been asked to modify a report (which unfortunately was written horribly!! not by me!) to include a count of days. Please note the "Days" is not calculated using "StartDate" & "EndDate" below. The problem is, there are multiple rows per record (users want to see the detail for start & enddate), so my total for "Days" are counting for each row. How can I get the total 1 time without the total in column repeating?
This is what the data looks like right now:
ID Description startdate enddate Days
REA145681 Emergency 11/17/2011 11/19/2011 49
REA145681 Emergency 12/6/2011 12/9/2011 49
REA145681 Emergency 12/10/2011 12/14/2011 49
REA146425 Emergency 11/23/2011 12/8/2011 54
REA146425 Emergency 12/9/2011 12/12/2011 54
I need this:
ID Description startdate enddate Days
REA145681 Emergency 11/17/2011 11/19/2011 49
REA145681 Emergency 12/6/2011 12/9/2011
REA145681 Emergency 12/10/2011 12/14/2011
REA146425 Emergency 11/23/2011 12/8/2011 54
REA146425 Emergency 12/9/2011 12/12/2011
Help please. This is how the users want to see the data.
Thanks in advance!
Liz
--- Here is the query simplified:
select id
,description
,startdate -- users want to see all start dates and enddates
,enddate
,days = datediff(d,Isnull(actualstardate,anticipatedstartdate) ,actualenddate)
from table
As you didn't provide the data of your tables I'll operate over your result as if it was a table. This will result in what you're looking for:
select *,
case row_number() over (partition by id order by id)
when 1 then days
end
from t
Edit:
Looks like you DID added some SQL code. This should be what you're looking for:
select *,
case row_number() over (partition by id order by id)
when 1 then
datediff(d,Isnull(actualstardate,anticipatedstartdate) ,actualenddate)
end
from t
That is a task for the reporting tool. You will have to write something like he next code in teh Display Properties of the Days field:
if RowNumber > 1 AND id = previous_row(id)
then -- hide the value of Days
Colour = BackgroundColour
Days = NULL
Days = ' '
Display = false
... (anything that works)
So they want the output to be exactly the same except that they don't want to see the days listed multiple times for each ID value? And they're quite happy to see the ID and Description repeatedly but the Days value annoys them?
That's not really an SQL question. SQL is about which rows, columns and derived values are supposed to be presented in what order and that part seems to be working fine.
Suppressing the redundant occurrences of the Days value is more a matter of using the right tool. I'm not up on the current tools but the last time I was, QMF was very good for this kind of thing. If a column was the basis for a control break, you could, in effect, select an option for that column that told it not to repeat the value of the control break repeatedly. That way, you could keep it from repeating ID, Description AND Days if that's what you wanted. But I don't know if people are still using QMF and I have no idea if you are. And unless the price has come way down, you don't want to go out and buy QMF just to suppress those redundant values.
Other tools might do the same kind of thing but I can't tell you which ones. Perhaps the tool you are using to do your reporting - Crystal Reports or whatever - has that feature. Or not. I think it was called Outlining in QMF but it may have a different name in your tool.
Now, if this report is being generated by an application program, that is a different kettle of Fish. An application could handle that quite nicely. But most people use end-user reporting tools to do this kind of thing to avoid the greater cost involved in writing programs.
We might be able to help further if you specify what tool you are using to generate this report.
It would seem that there is a much simpler way to state the problem. Please see Edit 2, following the sample table.
I have a number of different products on a production line. I have the date that each product entered production. Each product has two identifiers: item number and serial number I have the total number of labour hours for each product by item number and by serial number (i.e. I can tell you how many hours went into each object that was manufactured and what the average build time is for each kind of object).
I want to determine how (if) varying the length of production runs affects the average time it takes to build a product (item number). A production run is the sequential production of multiple serial numbers for a single item number. We have historical records going back several years with production runs varying in length from 1 to 30.
I think to achieve this, I need to be able to assign 'run id'. To me, that means building a query that sorts by start date and calculates a new unique value at each change in item number. If I knew how to do that, I could solve the rest of the problem on my own.
So that suggests a series of related questions:
Am I thinking about this the right way?
If I am on the right track, how do I generate those run id values? Calculate and store is an option, although I have a (misguided?) preference for direct queries. I know exactly how I would generate the run numbers in Excel, but I have a (misguided?) preference to do this in the database.
If I'm not on the right track, where might I find that track? :)
Edit:
Table structure (simplified) with sample data:
AutoID Item Serial StartDate Hours RunID (proposed calculation)
1 Legend 1234 2010-06-06 10 1
3 Legend 1235 2010-06-07 9 1
2 Legend 1237 2010-06-08 8 1
4 Apex 1236 2010-06-09 12 2
5 Apex 1240 2010-06-10 11 2
6 Legend 1239 2010-06-11 10 3
7 Legend 1238 2010-06-12 8 3
I have shown that start date, serial, and autoID are mutually unrelated. I have shown the expectation that labour goes down as the run length increases (but this is a 'fact' only via received wisdom, not data analysis). I have shown what I envision as the heart of the solution, that being a RunID that reflects sequential builds of a single item. I know that if I could get that runID, I could group by run to get counts, averages, totals, max, min, etc. In addition, I could do something like hours/ to get percentage change from the start of the run. At that point I could graph the trends associated with different run lengths either globally across all items or on a per item basis. (At least I think I could do all that. I might have to muck about a bit, but I think I could get it done.)
Edit 2: This problem would appear to be: how do I get the 'starting' member (earliest start date) of each run when I don't already have a runID? (The runID shown in the sample table does not exist and I was originally suggesting that being able to calculate runID was a potentially viable solution.)
AutoID Item
1 Legend
4 Apex
6 Legend
I'm assuming that having learned how to find the first member of each run that I would then be able to use what I've learned to find the last member of each run and then use those two results to get all other members of each run.
Edit 3: my version of a query that uses the AutoID of the first item in a run as the RunID for all units in a run. This was built entirely from samples and direction provided by Simon, who has the accepted answer. Using this as the basis for grouping by run, I can produce a variety of run statistics.
SELECT first_product_of_run.AutoID AS runID, run_sibling.AutoID AS itemID, run_sibling.Item, run_sibling.Serial, run_sibling.StartDate, run_sibling.Hours
FROM (SELECT first_of_run.AutoID, first_of_run.Item, first_of_run.Serial, first_of_run.StartDate, first_of_run.Hours
FROM dbo.production AS first_of_run LEFT OUTER JOIN
dbo.production AS earlier_in_run ON first_of_run.AutoID - 1 = earlier_in_run.AutoID AND
first_of_run.Item = earlier_in_run.Item
WHERE (earlier_in_run.AutoID IS NULL)) AS first_product_of_run LEFT OUTER JOIN
dbo.production AS run_sibling ON first_product_of_run.Item = run_sibling.Item AND first_product_of_run.AutoID run_sibling.AutoID AND
first_product_of_run.StartDate product_between.Item AND
first_product_of_run.StartDate
Could you describe your table structure some more? If the "date that each product entered production" is a full time stamp, or if there is a sequential identifier across products, you can write queries to identify the first and last products of a run. From that, you can assign IDs to or calculate the length of the runs.
Edit:
Once you've identified 1,4, and 6 as the start of a run, you can use this query to find the other IDs in the run:
select first_product_of_run.AutoID, run_sibling.AutoID
from first_product_of_run
left join production run_sibling on first_product_of_run.Item = run_sibling.Item
and first_product_of_run.AutoID <> run_sibling.AutoID
and first_product_of_run.StartDate < run_sibling.StartDate
left join production product_between on first_product_of_run.Item <> product_between.Item
and first_product_of_run.StartDate < product_between.StartDate
and product_between.StartDate < run_sibling.StartDate
where product_between.AutoID is null
first_product_of_run can be a temp table, table variable, or sub-query that you used to find the start of a run. The key is the where product_between.AutoID is null. That restricts the results to only pairs where no different items were produced between them.
Edit 2, here's how to get the first of each run:
select first_of_run.AutoID
from
(
select product.AutoID, product.Item, MAX(previous_product.StartDate) as PreviousDate
from production product
left join production previous_product on product.AutoID <> previous_product.AutoID
and product.StartDate > previous_product.StartDate
group by product.AutoID, product.Item
) first_of_run
left join production earlier_in_run
on first_of_run.PreviousDate = earlier_in_run.StartDate
and first_of_run.Item = earlier_in_run.Item
where earlier_in_run.AutoID is null
It's not pretty, and will break if StartDate is not unique. The query could be simplified by adding a sequential and unique identifier with no gaps. In fact, that step will probably be necessary if StartDate is not unique. Here's how it would look:
select first_of_run.AutoID
from production first_of_run
left join production earlier_in_run
on (first_of_run.Sequence - 1) = earlier_in_run.Sequence
and first_of_run.Item = earlier_in_run.Item
where earlier_in_run.AutoID is null
Using outer joins to find where things aren't still twists my brain, but it's a very powerful technique.