Is there a way to create a histogram based upon a single column of data? Say I have a column with values possible

a, b, c, d, e

. Is there a way to create a histogram where the x axis are the values

a, b, c, d, e

and the y axis are how many times each value occurred in the table?

Yes, that should be possible and quite simple. You will just need to write a MAQL metric, something like

SELECT COUNT(ID)

where

ID

is a primary key in your table/dataset. (In case of GD Cloud or GD CN m, you can also use

SELECT COUNT(dataset)

). This metric will give you a number of rows. And then use this metric in your chart together with the column. Note taht the column needs to be mapped to attribute in GoodData (not a fact). And you will get for each valie of the column number of rows where it appeared.

Oh okay. Well there's another wrinkle, what if I do a

SELECT COUNT (col1, col2)

, and then want to do a histogram of the calculated counts?

The two-parametric count in GoodData is a bit special. It does not give you count of the combination, but the first parameter is what you are counting (same as in single parametric COUNT) and the second parameter is a primary key of the dataset where you want to count it. For example COUNT(customer,customerID) would be number of customers in your list of customers. While COUNT(customer,orderID) would be number of customers in your orders tables (=those customers who have some order). This is assuming that you have dataset orders referencing the dataset customers in your logival data model.

Okay, yeah. So maybe I wasn't being as clear about what I needed as necessary. So, what I would like to do is group by a certain column (say

col1

) and then count the number of rows in those groups.

In GD Cloud you can calculate number of rows in dataset per col1 using

select count({dataset/dataset_id}) by col1

if you want to count number of col1 which have count in given range you can try following:

select count(col1) where (select count({dataset/dataset_id}) by col1) between ... and ...

… are the borders of intervals for the item in histogram. Then add multiple metrics with different ranges into your insight to create histogram.

🎉 New note created.

That worked! Thanks!