I am trying to retrieve a list of films from excel table and then in the lookup transformation find the missing genres in SQL table and add them. The result should be to add two new genres in SQL table which will automatically be given the id. Problem is that no match up lookup retrieves repeating genres.
Instead of adding just "software" and "the dark side" he add 3 time "software". Can i somehow do distinct in lookup?
As #MiguelH posted, using Sort "drop duplicate" facility solved my problem.
Related
I'm trying to join two tables together through a relationship but am having trouble.
My data is of a Building (type text) , which has Floors and Rooms which can both be a mix of text and number names (type any).
In order to create a relationship based on the Room column, I attempted to create a unqiue values table by referencing the Room_Mapping table, removing all by columns apart from Room, and removing duplicates.
However, when I try to join the two tables together based on the Room column, I keep getting the following message.
I have a feeling that this error is coming because the data type is a mix of text and numbers that it is getting confused because when I go to look at the data in Data view, I can see that the data for Phase and Room fields have been converted to Text type.
Please can any one help? I have attached a link to the workbook and data in the link below.
Room_Mapping Data and PBI Workbook
Many thanks in advance.
The Room1/ROOM1 Room2/ROOM2 are the cause of your problem. You can solve this by adding one extra step to your Query.
You do this by right-clicking the Room column and choose Transform > Capitalize Each Word.
The problem is that you have 4 very similar cells:
Room1
ROOM1
Room2
ROOM2
in original table and them are duplicated yet in the Room_UniqueValues Table.
Try to replace this values like this in the Room_UniqueValues Table:
And null by "null" in the original table, for example:
Tell me if this are ok please!
I have imported 2 tables from SQL Server to Power BI desktop in which each table contain one column having binary data. And these columns are used to create the relationship between that two tables.
When I proceed to create the relationship in Power BI these columns are not visible. I also tried the conversion of that binary column to text and try to create the relationship but it shows an error: Can´t create a relationship between two columns because one of the columns must have the unique value.
How can I create the relationship in Power BI desktop using Binary data columns?
As TDP stated, there has to be duplicate values, whether or not it is through an error in your data source. Please review the query editor and enable viewing column distribution and check to see if there are duplicate values in either of the columns you wish to use.
You will either need to modify your data source or make the changes within query editor directly to remove these duplicates.
It should not matter what data type you are using (binary/text) if there are duplicates. You will not be able to stop duplicates occurring just by changing the datatype.
EDIT: Oops, old thread. The answer still applies, please remove if unnecessary.
Can´t create a relationship between two columns
because one of the columns must have the unique value
This is your issue. One end of the relationship must have a distinct set of values.
Once you converted binary to text you would have been able to create the relationship if you had a distinct set on one side.
This isn't to do with binary being converted to text.
I am working with a list of bowling matches and want to update all matches for a specific week and a specific division.
By use of VBA, how do I change values in a column in a list for rows filtered for specific criteria on another column?
Something like that might be possible, but it's extremely difficult to asses without looking at the data. In SQL you have a database with fields and items, Excel tables can hold the data in a similar fashion, yet again as in SQL you need to see the relationship between the fields and items, to asses what can be done.
I was referencing this question How to export all data from table to an insertable sql format?
while looking for a way to create an insert statement for a single row from a table without having to manually write it since the table has many columns. In my case I simply followed the steps listed then performed a ctrl-f search in the resulting script for the record I wanted then copied and pasted that single line to another query window but this would be terrible if I had hundreds of millions of rows. Is there a way to get the same functionality but tell the script generator I only want rows where id = value? Is there a better way to do this using only the out of the box Microsoft tools?
There is no way to do this, but you can do it by using a temp table
Create a new table by inset into and select those records which you want to insert.
Create the script and change the table name by using find and replace.
finally drop that temporary table.
So my goal here is to have a single search field in an application that will be able to search multiple tables and return results.
For example, two of these tables are "performers" and "venues" and there are the following performers: "John Andrews","Andrew Smith","John Doe" and the following venues: "St. Andrew's Church","City Hall". Is there a way to somehow return the first two performers and the first venue for a search of "Andrew"?
My first thought was to somehow get all the tables aggregated into a single table with three columns; "SearchableText","ResultType","ResultID". The first column would contain whatever I want searched (e.g. Performer name), the second would say what is being shown (e.g. Performer) and the third would say the item's ID (note: all my tables have auto-incrementing primary keys for ease). The question for this idea is it possible to somehow do this dynamically or do I have to add code to have a table that automatically fills whenever a new row is updated/added/deleted from the performers and venues table (perhaps via trigger?).
My application is written in MSAccess (I know, I know, but I have no choice) on top of a SQL Server backend. I'd prefer this happen through MSAccess so I don't have to have a "searchme" table sitting on my SQL Server but any good result is acceptable :)
I think you are looking for the "union" sql keyword
I'd use full text indexing in SQL server, have a single table with your searchable text, and forign keys in your main tables that link to the search table. This way you can order your results by relevance.
I think you have a schema problem. Querying a UNION is almost always evidence of that (though not in all cases).
The question to me is:
What are you returning as your result?
If you find a person, are you displaying a list of people?
Or if you find a venue, a list of venues?
Or a mix of both?
I would say that if you want to return a list of both, then you'd want something like this:
SELECT tblPerson.PersonID, tblPerson.LastName & ", " & tblPerson.FirstName, "Person"
FROM tblPerson
WHERE tblPerson.LastName LIKE "Andrew*"
OR tblPerson.FirstName LIKE "Andrew*"
UNION
SELECT tblVenue.Venue, tblVenue.Venue, "Venue"
FROM tblVenue
WHERE tblVenue.Venue LIKE "Andrew*"
ORDER BY Venue
This will give a list of the matches indicating which is a person and which a venue, and allow you to then select one of those and open a detail view (by checking the value in the third field).
What you definitely don't want to do is this:
SELECT tblPerson.PersonID, tblPerson.LastName & ", " & tblPerson.FirstName, "Person"
FROM tblPerson
UNION
SELECT tblVenue.Venue, tblVenue.Venue, "Venue"
FROM tblVenue
then saving that and trying to query it on the 2nd column. That will be extremely inefficient. You want your WHERE clause to be on fields that can be searched via the index, and that means each subquery of your UNION needs to have an appropriate WHERE clause.