Skip to main content
Announcements
NEW: Seamless Public Data Sharing with Qlik's New Anonymous Access Capability: TELL ME MORE!
cancel
Showing results for 
Search instead for 
Did you mean: 
CK_WAKE
Creator
Creator

replace null to zero in table chart

Hi There, I have a table where Plant_no groups WORK_ORDER_KPI, HOT_JOINT_KPI, and K1VA_AGE_KPI, and summing them in the pivot table to replace null to zero does not give the correct values. All I want is to replace null with 0. Note: all the columns comes from diffrent table/ model.

Any hacks will be much appreciated. I have attached the data and screenshot of a table view to this post.

2 Solutions

Accepted Solutions
marksouzacosta

I think there is something wrong in your Load Script. These are my numbers with the Excel file you sent:

marksouzacosta_0-1721705857881.png

 

Read more at Data Voyagers - datavoyagers.net

View solution in original post

CK_WAKE
Creator
Creator
Author

i believe so, i will check the data model.

View solution in original post

8 Replies
SterreKapteijns
Partner - Creator
Partner - Creator

If you want to fill the empty values with zeros you could use a mapping table:

Map_Null:

mapping load

Null(), 'Unkown' 

AutoGenerate 1;

MAP K1VA_AGE_KPI using Map_Null;

 

[table name]:

load

EmptyIsNull(K1VA_AGE_KPI) as K1VA_AGE_KPI

etc. 

 

you could also use an if-statement, if(len(trim(K1VA_AGE_KPI)) = 0, 'NA', K1VA_AGE_KPI) as ...

 

BrunPierre
Partner - Master
Partner - Master

Hi, both ALT() and Coalesce() functions are used to handle null values or to provide alternative values when encountering nulls or missing data.

CK_WAKE
Creator
Creator
Author

Hi @SterreKapteijns ,  and @BrunPierre 

Thanks for getting back. Unfortunately, the code you provided is not working. That is because not all the plant_no (dimension) will have hot joint kpi attached. But if I remove the dimension and keep only the hot joint column, it appears correct.
 
I have attached the screenshot to show how the table looks when I remove the dimension column (Plant_no) without removing it.

I appreciate your guidance here.

marksouzacosta

What if you add these columns as Measures instead of Dimensions?

marksouzacosta_0-1721699377023.png

 

Read more at Data Voyagers - datavoyagers.net
CK_WAKE
Creator
Creator
Author

I tried to achieve this by summing it up. However, the summation gave me incorrect numbers. I randomly checked the number for Plant No: A801.4 DIS. It has a 50, 60, and 80 RISK SCORE , and it is correct; however, when I apply the summation it shows incorrect one. I hvae attached the sampel comparision below.  

marksouzacosta

I think there is something wrong in your Load Script. These are my numbers with the Excel file you sent:

marksouzacosta_0-1721705857881.png

 

Read more at Data Voyagers - datavoyagers.net
CK_WAKE
Creator
Creator
Author

i believe so, i will check the data model.

CK_WAKE
Creator
Creator
Author

The issue was in the data model. After revising the script, it is working fine. Thanks all for your time in addressing this.