How to count distinct values in a list

Question

I am fairly new to writing queries in Snowflake and have run into a hiccup. I am trying to count how many times an item appears in a list all in the same column.

I was able to use the flatten function and then tried to add in the count function with no luck.

Here is a dummy version of my data:

Ticket#              Tasks 
1               ["cut apple","peel orange","slice cheese"]
2               ["slice cheese","peel orange"]
3               ["cut apple"]
4               ["cut apple","slice cheese"]
5               ["cut apple", "chop kiwi"]

Here is what I want the output to look like: (hopefully auto populating the distinct list of tasks in desc order)

Tasks               Quantity
Cut Apple               4
Slice Cheese            3
Peel Orange             2
Chop Kiwi               1

Answer 1

Too long for a comment, but guidance for you to look into, then try to write a sample query. While you have the opportunity to do so while learning, I would look into Data Normalization and adjust your "Tasks" column.

You should have a secondary lookup table that has a primary key ID and a description of each unique task (you'll see in the data normalization). So you can follow along from your data context to the document, I will provide the layout examples and see how that helps you.

Starting with your lookup task table...

Tasks Table
TaskID   TaskDescription
1        cut apple
2        peel orange
3        slice cheese
4        chop kiwi

Then, you would have another table that has TicketID, and a third table shows multiple records for each TicketID.

Ticket Table
TicketID  ExPurchaseDate
1         someDate
2         sameDate
3         etc...

Now, a detail table per ticket.

TicketTasks Table
TicketTaskID  TicketID   TaskID
1             1          1
2             1          2
3             1          3
4             2          3
5             2          2
6             3          1
7             4          1
7             4          3
8             5          1
9             5          4

Try to digest this some with the normalization and then look into writing a sql query with COUNT(*) and GROUP BY. More than happy to help you more after, but hope this HELPs guide you some.

Answer 2

Step 1: Define a normalized data schema and put the schema into a database.

Your normalized data schema might look something like this:

Step 2: Add your data

Step 3: Then you will be able to use SQL COUNT with DISTINCT to find the unique rows in your data table(s)

SQL COUNT with DISTINCT