Unlock a world of possibilities! Login now and discover the exclusive benefits awaiting you.
Hi All,
I will start with my data model, I only have one table, Transactions, and the Master Calendar,
Transactions:
Load
event_id,
previous_event_id,
trans_id,
barcode,
item_id,
item_type,
location,
operator,
event_day,
Hour(timestamp) as hour,
timestamp
From
(qvd);
This table contains all the transactions from each event, every year the event might reoccur, but it is never on the same date, that is why I count each day of the event in the event_day field (1, 2, 3, 4...).
I want to create three charts: 1. dimension: hour, expression: count({<item_type={'P'}>}item_id)
2. dimension: location, expression: count({<item_type={'P'}>}item_id)
3. dimension: operator, expression: count({<item_type={'P'}>}item_id)
To give you an example:
2014: event_id = 258
2015: event_id = 374, previous_event_id = 258
I did get it to kind of work, I did a resident load from Transactions table:
Transactions_previous:
Load
event_id as previous_event_id,
trans_id as trans_id_prev,
barcode as barcode_prev,
item_id as item_id_prev,
item_type as item_type_prev,
location as location_prev,
operator as operator_prev,
event_day,
hour,
timestamp as timestamp_prev
Resident Transactions;
This creates a synthetic key:
$Syn 1 Table:
previous_event_id,
event_day,
hour;
I only created the hourly comparison and ran into an issue
Looks really good, but in 2014 there were transactions after 20:00 pm, so it only took the hours from the selected event_id.
this is the expression that I used for 2015: count({<item_type={'P'}>}item_id)
for 2014: count({<item_type_prev={'P'}>}item_id_prev)
Does anyone have any suggestions?
Message was edited by: Smaranda Fetila I added some sample data (Transactions.xlsx) and my application (Event_Transactions.qvw) and the second application showing what I tried doing with the resident load (Event_Transactions - Resident Load.qvw), any help would be appreciated! I don't have much experience with QlikView, if I did something wrong please let me know.
Message was edited by: Smaranda Fetila Also the end result is showed in the application with some images.
Can you provide some sample data to work with and the expected result???
You don't need this second resident load. The different years could you fetch per set analysis like:
count({< year = {2015}>} item_id)
count({< year = {2014}>} item_id)
or more dynamically
count({< year = {"$(=max(year))"}>} item_id) // current year based on selections
count({< year = {"$(=max(year)-1)"}>} item_id) // previous year
Further I suggest to remove the timestamp and split it into two fields like:
date(floor(timestamp)) as date,
time(frac(timestamp)) as time
and to use a Master-Time-Table - see here for more informations about date-fields:
How to use - Master-Calendar and Date-Values
- Marcus
Please provide sample data and desired output you want
Regards
Ankit Bisht
any reason for not renaming
event_day,
hour fields?
I think you only need to keep them the same if you want to join both the tables otherwise rename them . that shouuld get rid of the synthetic key issue and hopefully give you your results..
It will be really great, if you could attach a sample so that people could come up with a good solution to your issue
HTH
Sasi
Marcus,
I tried what you recommended, but when I select the event_id that I want to see it basically compares the data from this year to the same data. Don't I have to differentiate it somehow?
I attached some sample data and the application, could you take a look at it?
For the hourly comparison I need all three fields in common. Without the synthetic key it wouldn't compare the right data, would it? I attached some sample data and the app, could you take a look?
Thank you,
Smaranda
I added some sample data, the application and the end expected result in the application.
I added some sample data, the application and the end expected result in the application.
Actually I couldn't look into your examples but I think you have with your one table everything what you need to make your calculations. I suggest you have a look on what set analysis is and how it worked - it's really powerful:
Set Analysis: syntaxes, examples
Further you should definitely avoid synthetic keys - see why and how you get started with qlik datamodels:
Get started with developing qlik datamodels
- Marcus