I would like to receive your help in order to create an aggregate field in the Tableau tool.
I have a data source. The data source contains a few records (tuples/lines), and each record contains some attributes. I.e. a numeric value, a unique attribute (Key) and string.
For example, the first two rows share the same key (ID=A). This entity contains 2 records, the numeric value for each record is different (1 and 5).
I would like to create a new field. This new field should contain the minimum numeric value for each unique ID (entry).
In order to create this new field, I created a calculated field. The field’s formula is- min(NumericField).
The problem occurred when I was trying to use this calculated aggregated field aggregate in order to create chats. In this case, the aggregation is no longer performed on my key field.
For example, I would like to calculate the average for each category (The avg should be calculated for the aggregated new field)
Do you have any idea in order to resolve the mentioned problem?
Thank you! ☺️
An example
Try using LOD as below:
{fixed [id]: avg([newcalculatedfield])}
Let me know if you need more help. thanks!
Related
Assessing React-data-grid, I see the filter example puts a filter field on each column. Is it possible to implement a global filter field for the whole table?
Demo
http://adazzle.github.io/react-data-grid/docs/examples/column-filtering
In Filemaker Pro 12, I am trying to write a formula for a calculation field that will sum a field in a related table based on another field in that same related table. The normal Filemaker sum equation would look like this:
Sum (Assets::Asset Quantity)
However, I need to specify that only quantities that are related to a field named Asset Type with a value of "Building" will be used to filter the values in Asset Quantity that will be used in the sum.
There are a couple of ways that you could do this:
A new Calculated field
First, you could add a new Calculation field to your Assets table called, say, Building Quantity, with a Calculated Value of:
If (Asset Type = "Building" ; Asset Quantity ; 0)
And then you can use the sum of this new Building Quantity just like you were using Sum(Assets::Asset Quantity) before.
A new relationship
Second, you could add a new Calculated field to your main table with the value always equal to "Building" and then add a new table occurrence of the Assets table. We'll call it "BuildingAssets" and set the relationship so that your IDs match and also your new "Building" field matches the Asset Type
Summary ID \____________/ BuildingAssets::Summary ID
BuildingText / \ BuildingAssets::Asset Type
Then you will use
Sum (BuildingAssets::Asset Quantity)
instead of Sum (Assets::Asset Quantity) so that you only pull the Building types through.
ExecuteSQL
Finally, FileMaker 12 introduced the ExecuteSQL step. This may be the most elegant way to do the above because it doesn't involve changing any schema. The statement would be somethign like:
SELECT
SUM (Asset Quantity)
FROM
Assets
WHERE
Summary ID = ID AND
Asset Type = Building
For more information check out FileMaker's page: http://www.filemaker.com/12help/html/func_ref3.33.6.html
Also check out the FileMaker SQL Sugar ("#") Module for help building queries: http://www.modularfilemaker.org/2013/03/filemaker-sql-sugar/
I have a simple SSRS report that displays data from one table. What I want to do is have a distinct list from that table displayed in a drop down list for the user to select. If I only use one dataset I can get it to display, but it displays values from the column multiple times.
Example
Bob
Bob
Bob
Cathy
Cathy
If I create a second dataset that will list distinct values I get the following error message:
An Error occurred during local report processing. The definition of the report is invalid. The Variable expression for the report 'body' refers directly to the field without specifying a dataset aggregate. When the report contains multiple datasets, field references outside of a data region must be contained within aggregate functions which specify a dataset scope.
I"m trying to follow the example I found here:
http://msdn.microsoft.com/en-us/library/aa337400.aspx
The second dataset is only for the parameter list. I don't understand why it's causing problems with the actual report.
It's impossible to tell exactly where without the report definition, but there is an item on the report that is referencing a field or Dataset, and was implicitly using the only Dataset present in the report, but now doesn't know which Dataset to use once more than one is added to the report.
For example, when you create a table you can set a Dataset associated with it. If this is not set and there is only one Dataset, it doesn't matter as it will take the only one available. Once you add a new Dataset, the table doesn't know which one to use and you'll get the error you're seeing.
Another way to get the error is specifying a field in an expression, e.g. in a TextBox in the report somewhere without specifying the scope; just set the scope to a particular Dataset e.g. if you have:
=Count(Fields!name.Value)
change this to:
=Count(Fields!name.Value, "DatasetToUse")
If you've only got one Dataset the first expression will run fine by using the only one available, but once you add another it won't know which to use and it will error.
in the query (SQL) you should add DISTINCT clause at the beginning, that way, you will get only one record per value. Check out http://www.w3schools.com/sql/sql_distinct.asp
Double click the Dataset which contains that field.
Go to fields on the left and delete that field.
Add new field by clicking Add -> Query Field.
Just type in the name of the new field under field name and field source.
It happens when you have added a field by selecting "Calculated Field" instead of "Query Field" from Dataset Fields list tab.
Cheers,
Ahmed Latif
I would like to have the distinct values of a repeating field listed in another field (in browse mode). The case is as follows:
I have a field that contains country names. The country names in this field may repeat themselves, thus when using the "List" function I get something like "France, France, France, Germany, Germany, Hungary".
How can I create a field that lists all the values from my country field, but has it grouped as "France, Germany, Hungary"?
In the case I could directly use a SQL query to interfere with the FileMaker databse I would use the GROUP BY statement.
To make a summary of all the values across every record, do as follows:
Make a new value list labeled 'Countries' (File Menu > Manage > Value Lists)
Make the value list 'Use Values From Field' and specify your repeating field
Create a new Calculation field, 'Listed Countries'
Set the calculation to type 'Text' and with the following code:
ValueListItems ( Get(FileName) ; "Countries" )
If you'd like to find the value for only the current record:
Make a new Table Occurrence, 'NewTO' of the same base table and link the two records by a unique index.
Change the 'Countries' value list so that it obtains values from 'NewTO' and your repeating field.
Select 'Only include related values starting from' and select your original Table Occurrence
If you'd like the list to update as the repeating field value changes, make certain that you Do Not Store Calculation Results for the field.
I have a table like this:
Application,Program,UsedObject
It can have data like this:
A,P1,ZZ
A,P1,BB
A,P2,CC
B,F1,KK
I'd like to create a layout to show:
Application,# of Programs
A,2
B,1
The point is to count the distinct programs.
For the life of me I can't make this work in FileMaker. I've created a summary field to count programs resetting after each group, but because it doesn't eliminate the duplicate programs I get:
A,3
B,1
Any help much appreciated.
Create a a summary field as:
cntApplicaiton = Count of Application
Do this by going into define fields, create a field called cntApplication, type summary. In the options dialogue make the summary field a count on application
Now create a new layout with a subsummary part and nobody. The subsummary should be sorted on Application. Put the Application and cntApplication fields in subsummary. If you enter browse mode and sort by Application you ought to get the data you want.
You can also create a calc field with the formula
GetSummary(cntApplication; Application)
This will allow you to use the total number of Applications with in a record
Since I also generate the data in this form, the solution I've adopted is to fill two tables in FileMaker. One provides the summary view, the other the detailed view.
I think that your problem is down to dupliate records and an inadequate key.
Create a text field called "App_Prog". In the options box set it to an auto-enter calc, unchecking the 'Do not replace...' option, and use the following calc:
Application & "_" & Program
Now create a self join to the table using App_Prog as the field on both sides, and call this 'MatchingApps'.
Now, create (if you don't alread have one) a unique serial number field, 'Counter' say, and make sure that you enter a value in each record. (Find all, click in the field, and use serial number option in'Replace Field Contents...')
Now add a new calc field - Is_Duplicate with the following calc...
If (Counter = MatchingApps::Counter; "Master Record" ; "Duplicate")
Finally, find all, click in the 'Application field, and use 'Replace Field Contents...' with a calculation to force the auto-enter calc for 'App_Prog' to come up with a value.
Where does this get you? You should now have a set of records that are marker either "Master Record" or "Duplicate". Do a find on "Master Record", and then you can perform your summary (by Application) to do a count of distinct application-program pairs.
If you have access to custom functions (you need FileMaker Pro Advanced), I'd do it like this:
Add the RemoveDuplicates function as found here (this is a recursive function that takes a list of strings and returns a list of unique values).
In the relationships graph, add another occurrence of your table and add an Application = Application relationship.
Create a calculated field in the table with the calculation looking something like this:
ValueCount(RemoveDuplicates(List(TABLE2::Program)))
You'll find that each record will contain the number of distinct programs for the given application. Showing a summary for each application should be relatively trivial from here.
I think the best way to do this is to create a separate applications table. So as you've given the data, it would have two records, one for A and one for B.
So, with the addition of an Applications table and your existing table, which I'll call Objects, create a relationship from Applications to Objects (with a table occurrence called ObjectsParent) based on the ApplicationName as the match field. Create a self join relationship between Objects and itself with both Application and Program as the match fields. I'll call one of the "table occurrences" ObjectsParent and the other ObjectsChildren. Make sure that there's a primary key field in Objects that is set to auto-enter a serial number or some other method to ensure uniqueness. I'll call this ID.
So your relationship graph has three table occurrences:
Applications::Applicaiton = ObjectsParent::Application
ObjectsParent::Application = ObjectsChildren::Application, ObjectsParent::Program = ObjectsChildren::Program
Now create a calculation field in Objects, and calculating from the context of ObjectsParent, give it the following formula:
AppCount = Count( ObjectsChildren::ID )
Create a calculation field in Applications and calculating from the context of the table occurrence you used to relate it to ObjectsParent with the following formula:
AppCount = ObjectsParent::AppCount
The count field in Objects will have the same value for every object with the same application, so it doesn't matter which one you get this data from.
If you now view the data in Applications in list view, you can place the Applications::Application and Applications::AppCount fields on the layout and you should get what you've requested.