Skip to main content

Qlik

cancel
Turn on suggestions
Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.
Showing results for 
Search instead for 
Did you mean: 
Announcements
Join us at Qlik Connect for 3 magical days of learning, networking,and inspiration! REGISTER TODAY and save!
cancel
Turn on suggestions
Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.
Showing results for 
Search instead for 
Did you mean: 
SK28
Creator
Creator
‎2023-11-15 05:33 AM

find ORG names which has Skews in line chart qliksense

Hi,

 

I need find the ORG name which has sudden upwards Skews  in the line chart, based on the count of the tickets, there are more than 400+ ORGs and the date range is Consistent across for all ORGs, I need to find the ORGs which has sudden SPIKE in the tickets

Here is the Data Sample data:

ORG date Count of tickts
TCS 10/11/2023 10
TCS 10/12/2023 12
TCS 10/13/2023 30
TCS 10/14/2023 12
TCS 10/15/2023 10
CTS 10/11/2023 10
CTS 10/12/2023 25
CTS 10/13/2023 9
CTS 10/14/2023 12
Wipro 10/14/2023 8
Wipro 10/15/2023 7
Infy 10/11/2023 6
Infy 10/12/2023 5

 

 

here is the line chart for above table

SK28_2-1700044122092.png

 


required output will be --> TCS on 10/13/2023 and CTS on 10/12/2023 
or 
TCS,CTS ORG names will also works

2nd Requirement:

Find the Name of the ORG, which has increase in the Count of Tickets

ORG Name Count of tickts
TCS 10/11/2023 10
TCS 10/12/2023 12
TCS 10/13/2023 13
TCS 10/14/2023 14
TCS 10/15/2023 17
TCS 10/16/2023 18
TCS 10/17/2023 19
TCS 10/18/2023 20
CTS 10/11/2023 11
CTS 10/12/2023 11
CTS 10/13/2023 12
CTS 10/14/2023 10
CTS 10/15/2023 8
CTS 10/16/2023

7

 

SK28_3-1700044282918.png

 

Required Output will be - TCS

Appreciate your help!

667 Views
6 Replies
WaltShpuntoff
Employee
Employee
‎2023-11-16 08:15 AM

I understand you are trying to identify the "spiked" Org here. The question I have is where you are looking to see that information.

Keep in mind that you can select the date with the spike which will narrow all activity down to that date. You can define alternate dimensions and measures in the chart, so if you define an alternate dimension for the Org you could then flip the display and I believe that will show you what you are looking for.

- ws

627 Views
0 Likes
Or
MVP
MVP
‎2023-11-16 08:24 AM

This isn't a particularly clear requirement... "Spike" and "increase" both need to be quantified.

Spikes should be fairly easy to find once defined properly - either by comparing the count of tickets to the Previous() value, or by comparing the count to a count from the day before using set analysis. You would then check if that value is higher than whatever threshold you've defined,e.g. if the ratio is greater than 200%.

A gradual increase is a bit trickier, but again this depends on the exact definition. If you're just looking for cases where the total for today is more than, say, 50% of the total from last week, regardless of behavior in the middle section, that would be the same as the solution for a spike. If it needs to be a full rising line, where each date is higher than the one before, I can't think of an easy way to do it in the front end without manually comparing the required number of days. Perhaps someone else will have a way to do that.

625 Views
0 Likes
SK28
Creator
Creator
‎2023-11-16 02:20 PM
Author

@SK28 wrote:

Hi,

 

I need find the ORG name which has sudden upwards Skews  in the line chart, based on the count of the tickets, there are more than 400+ ORGs and the date range is Consistent across for all ORGs, I need to find the ORGs which has sudden SPIKE in the tickets

Here is the Data Sample data:

ORG date Count of tickts
TCS 10/11/2023 10
TCS 10/12/2023 12
TCS 10/13/2023 30
TCS 10/14/2023 12
TCS 10/15/2023 10
CTS 10/11/2023 10
CTS 10/12/2023 25
CTS 10/13/2023 9
CTS 10/14/2023 12
Wipro 10/14/2023 8
Wipro 10/15/2023 7
Infy 10/11/2023 6
Infy 10/12/2023 5

 

 

here is the line chart for above table

SK28_2-1700044122092.png

 


required output will be --> TCS on 10/13/2023 and CTS on 10/12/2023 
or 
TCS,CTS ORG names will also works

2nd Requirement:

Find the Name of the ORG, which has increase in the Count of Tickets

ORG Name Count of tickts
TCS 10/11/2023 10
TCS 10/12/2023 12
TCS 10/13/2023 13
TCS 10/14/2023 14
TCS 10/15/2023 17
TCS 10/16/2023 18
TCS 10/17/2023 19
TCS 10/18/2023 20
CTS 10/11/2023 11
CTS 10/12/2023 11
CTS 10/13/2023 12
CTS 10/14/2023 10
CTS 10/15/2023 8
CTS 10/16/2023

7

 

SK28_3-1700044282918.png

 

Required Output will be - TCS

Appreciate your help!

@SK28 wrote:

Hi,

 

I need find the ORG name which has sudden upwards Skews  in the line chart, based on the count of the tickets, there are more than 400+ ORGs and the date range is Consistent across for all ORGs, I need to find the ORGs which has sudden SPIKE in the tickets

Here is the Data Sample data:

ORG date Count of tickts
TCS 10/11/2023 10
TCS 10/12/2023 12
TCS 10/13/2023 30
TCS 10/14/2023 12
TCS 10/15/2023 10
CTS 10/11/2023 10
CTS 10/12/2023 25
CTS 10/13/2023 9
CTS 10/14/2023 12
Wipro 10/14/2023 8
Wipro 10/15/2023 7
Infy 10/11/2023 6
Infy 10/12/2023 5

 

 

here is the line chart for above table

SK28_2-1700044122092.png

 


required output will be --> TCS on 10/13/2023 and CTS on 10/12/2023 
or 
TCS,CTS ORG names will also works

2nd Requirement:

Find the Name of the ORG, which has increase in the Count of Tickets

ORG Name Count of tickts
TCS 10/11/2023 10
TCS 10/12/2023 12
TCS 10/13/2023 13
TCS 10/14/2023 14
TCS 10/15/2023 17
TCS 10/16/2023 18
TCS 10/17/2023 19
TCS 10/18/2023 20
CTS 10/11/2023 11
CTS 10/12/2023 11
CTS 10/13/2023 12
CTS 10/14/2023 10
CTS 10/15/2023 8
CTS 10/16/2023

7

 

SK28_3-1700044282918.png

 

Required Output will be - TCS

Appreciate your help!



Thanks for the response @Or & @WaltShpuntoff 

 

The whole intention here is We want to find these two observations and we are trying to automate them , trigger them to LEADERSHIP team so that...using NPRINTING

then users will start using the dashboard more intensively to take corrective decisions

Doing this help

1)Help users identify these two observations and act on them

2) Qliksense product usuage will Improve in our Organization.

for request 1)

and we don't have any thesholds limits set here, if there is any sudden spike let's consider say >=75%  

we need to find the ORG name...! which has sudden spike or skew in the data 

and for point 2) if the count of tickets are continuously growing by Date... we need to find those observations as well.

we tried it using SKEW & Stdev But I got struck in midway...!! using above data even counts which are falling are eventually coming down continuously are also showing up.

 

If you can please help me with the logic, That would be a great help...!! like Previous() it will be helpful 

@Or  Spikes should be fairly easy to find once defined properly - 

either by comparing the count of tickets to the Previous() value, or by comparing the count to a count from the day before using set analysis. 

 

Using Above same Sample data please...! 

615 Views
0 Likes
WaltShpuntoff
Employee
Employee
‎2023-11-16 09:21 PM

I do not know your environment, but if you have the Insight Advisor active, you can ask it for Anomaly detection.

WaltShpuntoff_0-1700187655468.png

 

600 Views
Or
MVP
MVP
‎2023-11-19 06:23 AM

In response to SK28

Checking for a daily increase of 75% is pretty easy. You just pull today's value and yesterday's value and compare them, so your table would, with null/zero hidden, be a dimension of org, and a measure along the lines of:

if(Count({< Date = {"=Today()"} Tickets) / Count({< Date = {"=Today()-1"} Tickets) > 1.75, 'Spike')

575 Views
Aasir
Creator III
Creator III
‎2023-11-19 08:23 AM

In response to Or

Use below two expressions 

if(Count([Count of tickets]) > 0,
((Count([Count of tickets]) - Previous([Count of tickets])) / Previous([Count of tickets])) * 100,
null()
)

sum(if(Count([Count of tickets]) > 0,
((Count([Count of tickets]) - Previous([Count of tickets])) / Previous([Count of tickets])) * 100,
null()
))

567 Views
0 Likes