To GoodData homepage

Replies posted by JeffC

6 Replies

JeffC
JeffCParticipating Frequently
asked in Metrics and MAQL

Date Time - Date Time

I’ll review this.  Is there any plan to simplify this for customers with any upcoming enhancement.  I’m working towards a reference table to intervals in the day so this makes sense on the surface and will review this for shorter term.Also, wondering if you have any purpose use-case documentation - some of your customers have to be call center and I'm sure they’ve tackled the same thing we’re looking to do.  Any knowledge sharing here?

3
Radek Nováček
GoodData Employee
1 year ago
JeffC
JeffCParticipating Frequently
asked in Metrics and MAQL

Rollup/ Aggregation Wrong

Hi Jeff, could you please make our support user an admin in the account so we can verify a few things? this is done. It seems that the default behavior due to relational nature of goodata is to return a count distinct, but hoping i’m wrong or there is an easy solution here. :)COUNT(call_sid) should return 3, versus 2 which is the current behavior.calls|id   | length | call_sid||---------------------------||  1  |    11     |   X00  ||  2  |    22     |   X00  ||  3  |    40     |   A12  |

8
Joseph Heun
GoodData Employee
1 year ago
JeffC
JeffCParticipating Frequently
asked in Metrics and MAQL

Rollup/ Aggregation Wrong

Hi Jeff,  This is Joseph from the GoodData Support Team, and I’ve been looking into this with Julius as well. I’ve been looking into the insight trying to recreate it, but I see everything working as expected. The sums are correct, but when you use count it will count all available values. Therefore, the easiest thing to do would be to filter out the unwanted values within the insight itself; e.g. filter out the duplicated values you see.      My issue is a bit different.  The metric is producing 135 versus the expected 142.  I want the raw metric to produce the count of call_sid which is 142, but it’s incorrectly returning 135 unique rows.

8
Joseph Heun
GoodData Employee
1 year ago
JeffC
JeffCParticipating Frequently
asked in Metrics and MAQL

Rollup/ Aggregation Wrong

Hello Jeff,   The rollup is an underlying calculation which acts as a smart aggregation and it is calculated as if no attributes had been added to the report. It counts values that have missing references. Whereas the Sum counts what is actually being displayed in the report. That is the reason why Rollup might display a different result to Sum. We have a documentation describing aggregating function in more details here: https://help.gooddata.com/classic/en/dashboards-and-insights/reports/working-with-tables/aggregate-table-data   Regarding your question, it is really hard to evaluate this without knowing your actual LDM and relations within the datasets. I can see you have provided hyperlinks for the particular attributes. However, we do not have an access into your workspace. Therefore, if you still need further assistance, could you please invite our support user support+novelcx@gooddata.com? I’ve provided access to support+novelcx@gooddata.com

8
Joseph Heun
GoodData Employee
1 year ago
JeffC
JeffCParticipating Frequently
asked in Metrics and MAQL

Rollup/ Aggregation Wrong

I realize that even though call_sid in my table is not a primary key, it contained duplicate text strings.  This was not intentional, so we’ll fix the plumbing upstream.  But it does pose the question for counting other columns where there is duplicative data. How would I have built the metric if I wanted to calculate every matching row, including those with duplicate strings?

8
Joseph Heun
GoodData Employee
1 year ago
JeffC
JeffCParticipating Frequently
asked in Data models

Updating Data Models when Souece Updated

Thanks, to be clear after reading the guide, the only time I have a guided approach to building my model is during the initial creation where I can map the various columns, after the model is made, I have not manually add every new column to a table. Gooddata does not cross reference my copy of the data model with the source, nor does it volunteer that my model is now out of date due to it missing N columns offering any type of workflow to close the gap. I ask because this means that our resource needs to monitor all in scope tables and ensure we have tight controls to start.  While important, it’s a different expectation than a system that adds this missing value.

3
Iva Gasparekova
GoodData Employee
1 year ago
Welcome to the GoodData Community!Cookie settings