I have 3 tables, Pizza, Ingredient, and PizzaIngredients which is a join table between my first two tables.
Pizza
id
name
1
Margarita
2
Pepperoni
3
Vegetarian
Ingredient
id
name
1
Tomato
2
Pepperoni
3
Lettuce
PizzaIngredients
pizzaId
ingredientId
1
1
2
1
2
2
3
1
3
3
I have a list of ingredients, and I want to list all the pizza I can make with the ingredients I have. To be able to make a pizza, i need to have all its ingredients
If in my list of ingredients I have : [1], I want to return the pizza with id 1 since I can only make the Margarita pizza.
If in my list of ingredients I have : [1, 3], I want to return the pizza with id 1 and 3 since I can potentially make the Margarita or the Vegetarian.
What I've tried :
SELECT
pi.pizzaId
FROM
pizzaIngredients pi
WHERE
pi.ingredientsId NOT IN (
SELECT
DISTINCT pi2.ingredientsId
FROM
pizzaIngredients pi2
WHERE
pi2.ingredientsId NOT IN ('1')
)
But It also returns me pizza 2 and 3 which I dont want since I don't have all the ingredients. I use a postgresql database
try this :
SELECT pi.pizzaId
FROM pizzaIngredients pi
GROUP BY pi.pizzaId
HAVING array_agg(ingredientId) <# array [1,3]
dbfiddle
Related
I am trying to have all components that are part of a product appear on the same row as each other
I have two tables
PRODUCT
ID PRODUCTNUMBER DESCRIPTION TYPE STATUS KIT
1 (PK) 121 1 Apples and 1 Oranges FRUIT PACK YES Y
2 122 2 Brocolli & 2 Carrots VEG PACK NO Y
3 123 3 Strawberries and 3 Blueberries and 1 Pear FRUIT PACK YES Y
4 124 2 Plums and 1 Pears FRUIT PACK YES Y
5 125 4 Grapes and 2 Cabbage COMBO PACK YES Y
6 126 Apple FRUIT YES N
7 127 Orange FRUIT YES N
8 128 Pear FRUIT YES N
9 129 Onion VEG NO N
10 130 Blueberry FRUIT YES N
11 131 Strawberry FRUIT YES N
12 132 Plum FRUIT YES N
PRODUCTCOMPONENT
PRODUCT QTY
5 55
6 45
7 21
8 12
9 0
10 20
11 25
12 50
My SQL query should return:
SKU Description COMPONENT1 QTY1 COMPONENT2 QTY2 COMPONENT3 QTY3
121 1 Apples and 1 Oranges Apple 55 Orange 45
123 3 Strawberries and 3 Blueberries and 1 Pear Strawberries 25 Blueberry 20 Pear 12
124 2 Plums and 1 Pears Plum 50 Pear 12
I tried:
SELECT
PRODUCT.CODE, PRODUCT.DESCRIPTION,
PRODUCTCOMPONENT.PRODUCT, PRODUCTCOMPONENT.QTY
FROM
PRODUCT
INNER JOIN
PRODUCTCOMPONENT ON PRODUCTCOMPONENT.PRODUCT = PRODUCT.ID
WHERE
PRODUCT.STATUS = YES
AND PRODUCT.KIT = Y;
Any help would be appreciated
Okay, this is from memory, but I've verified the syntax with SQL Fiddle.
You are right that you need to start with PRODUCT and PRODUCTCOMPONENT. And the code you posted will give you the data you want -- but it won't have the answers in columns, just in rows.
So what you have is what I call a "rotation problem". You want to "swing" the data over 90 degrees (so to speak), and have multiple columns where you had multiple rows.
There is no automatic, built-in way to do this. But there are indirect ways.
What you're going to have to do is left-outer-join PRODUCTCOMPONENT to PRODUCT once for every set of columns you want to display component information.
If you have 2 columns, you'd need to do it twice. Since your max is 5, you'll need to do it 5 times.
This is why I asked how many components you could have per item. If you had an indefinite number, you'd be out of luck, because there just isn't a simple way to
automatically extend columns out to the right for as many sets of rows you happen to have. You have to do a new left join clause for every additional possible component!
Here's an example of the 2-column case, which should show you how to do the 5-column case:
-- In order to join to just the records from Row 1, we need to number them!
-- We'll do that in a CTE (Common Table Expression).
;
WITH Components as (
-- I don't know all the columns in PRODUCTCOMPONENT, but you presumably have a
-- parent and child ID. Substitute the true names of the columns for the
-- column names I'm using
SELECT ParentId
, ChildId
, Product -- I am assuming this is the product name
, Qty
-- The following line will assign a line number to each component within
-- a product. If there's a particular order you want the columns to appear in,
-- change the "Order by" part of the ROW_NUMBER() OVER expression.
, RowNumber = ROW_NUMBER() OVER (Partition By ParentId Order By ChildId)
FROM ProductComponent
)
SELECT Product.PRODUCTNUMBER as Code
, Product.DESCRIPTION
, Component1.Product as Component1
, Component1.Qty as Qty1
, Component2.Product as Component2
, Component2.Qty as Qty2
FROM Product
-- Note that since some products will have more components than others,
-- you need to left-join to the Components CTE to make sure that rows are
-- still returned even when they only have nulls.
LEFT OUTER JOIN Components as Component1
ON Product.ID = Component1.ParentID
AND Component1.RowNumber = 1
-- The second clause of the JOIN means that you'll only get rows back
-- from the CTE if the RowNumber assigned in the CTE is (in this case) 2.
LEFT OUTER JOIN Components as Component2
ON Product.ID = Component2.ParentID
AND Component2.RowNumber = 2
WHERE Product.STATUS = 'YES'
AND Product.KIT = 'Y';
I have a table filled with cars, another table that has the car IDs and the feature IDs and another table with the feature IDs and feature names.
A car can have multiple features in that table. The result of the query is, that i get multiple instances of the same car, each having a different feature.
I would like to have only one instance of the car (by ID) and combine all features into one column (list or array). I could do it in code after the query but i would like to do it in a query instead.
Cars table Car property table Property table
ID Name Car_ID property_ID property_id Property_name
1 Audi 1 1 1 Aircon
2 BMW 1 2 2 Autopilot
3 Mercedes 2 1
3 2
Result is:
1 Audi Aircon
1 Audi Autopilot
2 BMW Autopilot
3 Mercedes None
Result should be:
1 Audi [Aircon, Autopilot]
2 BMW Autopilot
3 Mercedes None
This seems like a basic aggregation query with joins:
select c.name, group_concat(Property_name) as properties
from cars c left join
car_properties cp
on c.car_id = cp.car_id left join
properties p
on p.property_id = cp.property_id
group by c.name
I have a table which updates every few hours i.e. new rows are added and data for existing rows may also change.
I am looking for an output where the new rows are added but the existing data does not get over-written with the changes but changes are show in a new column
I am using big query so some standard functions may not work. does this require loops?
Base table at T=0
id food
1 cake
2 pepsi
3 peanut
4 chicken
Base table at T=1 (a new row has been added i.e. id 5)
id food
1 cake
2 pepsi
3 peanut
4 chicken
5 watermelon
Output at T=1
id food change
1 cake NULL
2 pepsi NULL
3 peanut NULL
4 chicken NULL
5 watermelon NULL
Base table at T=2 (a new row has been added i.e. id-6 and food names for id-3 and id-5 have been changed)
id food
1 cake
2 pepsi
3 sushi
4 chicken
5 wrap
6 Cherry
Output at T=2
id food change
1 cake NULL
2 pepsi NULL
3 peanut sushi
4 chicken NULL
5 watermelon wrap
6 Cherry NULL
If you are using update to change the value, then there is nothing you can do. You have changed the data and you don't have a record of earlier data.
If you are inserting the data with a timestamp, then you can construct a query. Your data would look like:
id food timestamp
1 cake 0
2 pepsi 0
3 peanut 0
4 chicken 0
1 cake 1
2 pepsi 1
3 peanut 1
4 chicken 1
5 watermelon 1
Then a typical method to get the current values might be:
select id, food, prev_food
from (select t.id, t.food,
lag(t.food) over (partition by t.id order by t.timestamp) as prev_food,
dense_rank() over (order by t.timestamp desc) as seqnum
from t
) t
where seqnum = 1;
In BigQuery, this can actually be simplified to:
select id,
array_agg(food order by timestamp desc limit 1)[safe_ordinal(1)],
array_agg(food order by timestamp desc limit 2)[safe_ordinal(2)]
from t
group by id;
This does, however, assume that id do not disappear.
With the above input/output and under the assumption that the value can changes only once, in postgreSQL we could do something like this.
INSERT INTO foo (id, food)
VALUES
(
3,
'Sushi'
)
ON CONFLICT (id)
DO
UPDATE
SET change = excluded.food;
We could use similar methodology in MySQL but the syntax would be different.
I have 2 Tables in Access and I am trying to build a LookUp Query. (I am new to SQL)
Dogs
ID
DogName
Type (int) Either 1,2,3,4
ClassResults
ID
ClassEntered (int) 1-24
DogName
So in my lookup I am trying to find all the dogs from the DOGS table that if the ClassEntered is less than 12 select all the dogs with a TYPE 1 or 2.
Dogs Data Sample:
ID DogName Type
0 AAA 1
1 BBB 3
2 CCC 1
3 DDD 2
4 EEE 4
ClassResults Data Sample:
ID ClassEntered DogName
0 6 ?????
So, the Drop Downlist for the DogName should be Showing:
0, AAA, 1
2, CCC, 1
3, DDD, 2
SELECT DISTINCT Dogs.DogName FROM Dogs, ClassResults
WHERE (IIf([ClassResults].[ClassEntered] < 10,[Dogs.Type]<3,[Dogs.Type]>2)) ORDER BY Dogs.DogName;
SELECT DISTINCT Dogs.DogName FROM Dogs, ClassResults
WHERE (IIf([ClassResults].[ClassEntered] < 10,[Dogs.Type]<3,[Dogs.Type]>2)) ORDER BY Dogs.DogName;
I hope that makes sense.
Stephan
It's better to use joins:
select * from dogs d
left join ClassResults c on d.dogname=c.dogname
where c. ClassEntered<12 and d.type in (1,2)
If you want to pick data from two different tables, you have to utilize joins.
Is there a built-in way in Oracle 11 to check correlation of values in a varchar2 field? For example, given a simple table such as this:
MEAL_NUM INGREDIENT
--------------------
1 BEEF
1 CHEESE
1 PASTA
2 CHEESE
2 PASTA
2 FISH
3 CHEESE
3 CHICKEN
I want to get a numerical indication that based on MEAL_NUM, CHEESE is paired mostly with PASTA and to lessening degrees with BEEF, CHICKEN, and FISH.
My first inclination is to use the CORR function and transform the strings into a number perhaps by either enumerating them beforehand or grabbing the rownum from a unique select.
Any suggestions how to go about this?
You won't want to use CORR -- if you create a "food number" and assign Beef = 1, Chicken = 2, and Pasta = 3, then a correlation coefficient will tell you whether increased cheese correlates with increased "food number." But the "food number" being higher or lower doesn't mean anything since you made it up. So, don't use CORR unless your foods are actually ordered in some way, like numbers are.
The way statisticians talk about this is with levels of measurement. In the language of the linked article, MEAL_NUM is a nominal measure -- or maybe an ordinal measure if the meals happened in order, but either way, it's a really bad idea to use correlation coefficients on it.
You'll probably instead want to find something like "what percentage of Beef meals also have Cheese?" The following will return, for each ingredient, the number of meals containing it and also the number of meals containing it AND cheese. The trick is that COUNT only counts non-null values.
SELECT Other.Ingredient,
COUNT(*) AS TotalMeals,
COUNT(Cheese.Ingredient) AS CheesyMeals
FROM table Other
LEFT JOIN table Cheese
ON (Cheese.Ingredient = 'Cheese'
AND Cheese.Meal_Num = Other.Meal_Num)
GROUP BY Other.Ingredient
Warning: returns wrong results if you include an ingredient twice in any one meal.
Edit: It turns out you aren't interested in Cheese specifically. You really want all the pairs of "correlations." So, we can abstract "Cheese" out and call them just the First and Second ingredients. I've added a "PossibleScore" to this one which tries to act like a percentage-of-meals but doesn't give a strong score if there are very few instances of the ingredient.
SELECT First.Ingredient,
Second.Ingredient,
COUNT(*) AS MealsWithFirst,
COUNT(First.Ingredient) AS MealsWithBoth,
COUNT(First.Ingredient) / (COUNT(*) + 3) AS PossibleScore,
FROM table First
LEFT JOIN table Second
ON (First.Meal_Num = Second.Meal_Num)
GROUP BY First.Ingredient, Second.Ingredient
When sorted by score, this should return
PASTA CHEESE 2 2 0.400
CHEESE PASTA 3 2 0.333
BEEF CHEESE 1 1 0.250
BEEF PASTA 1 1 0.250
FISH CHEESE 1 1 0.250
FISH PASTA 1 1 0.250
CHICKEN CHEESE 1 1 0.250
PASTA BEEF 2 1 0.200
PASTA FISH 2 1 0.200
CHEESE BEEF 3 1 0.167
CHEESE FISH 3 1 0.167
CHEESE CHICKEN 3 1 0.167
Do a self join to get all the in ingredient combinations, then corr by the two meal_nums
SELECT t1.INGREDIENT, t2.INGREDIENT, CORR(t1.MEAL_NUM, t2.MEAL_NUM)
FROM TheTable t1, TheTable t2
WHERE t1.INGREDIENT < t2.INGREDIENT
GROUP BY t1.INGREDIENT, t2.INGREDIENT
Should give you something like:
BEEF CHEESE 0.999
BEEF PASTA 0.998
CHEESE PASTA 0.977
UPDATE: as Chris points out, this won't work as is. What I was hoping is that there might be some way to fudge a mapping from the ordinal meal_num to an interval (#Chris, thanks for the link) value. That may not be possible, in which case this answer wouldn't help.
Try DBMS_FREQUENT_ITEMSET:
--Create sample data
create table meals(meal_num number, ingredient varchar2(10));
insert into meals
select 1, 'BEEF' from dual union all
select 1, 'CHEESE' from dual union all
select 1, 'PASTA' from dual union all
select 2, 'CHEESE' from dual union all
select 2, 'PASTA' from dual union all
select 2, 'FISH' from dual union all
select 3, 'CHEESE' from dual union all
select 3, 'CHICKEN' from dual;
commit;
--Create nested table type to hold results
CREATE OR REPLACE TYPE fi_varchar_nt AS TABLE OF VARCHAR2(10);
/
--Find the items most frequently combined with CHEESE.
select bt.setid, nt.column_value, support occurances_of_itemset
,length, total_tranx
from
(
select
cast(itemset as fi_varchar_nt) itemset, rownum setid
,support, length, total_tranx
from table(dbms_frequent_itemset.fi_transactional(
tranx_cursor => cursor(select meal_num, ingredient from meals),
support_threshold => 0,
itemset_length_min => 2,
itemset_length_max => 2,
including_items => cursor(select 'CHEESE' from dual),
excluding_items => null))
) bt,
table(bt.itemset) nt
where column_value <> 'CHEESE'
order by 3 desc;
SETID COLUMN_VAL OCCURANCES_OF_ITEMSET LENGTH TOTAL_TRANX
---------- ---------- --------------------- ---------- -----------
4 PASTA 2 2 3
3 FISH 1 2 3
1 BEEF 1 2 3
2 CHICKEN 1 2 3
what about a query like that?
select t1.INGREDIENT, count(*)a
from table t1,
(select meal_num
from table
where INGREDIENT = 'CHEESE') t2
where t1.INGREDIENT <> 'CHEESE'
and t1.meal_num=t2.mealnum
group by t1.INGREDIENT;
the result should be the number of time each ingredient share a meal_num with CHEESE.