I have a table with 59 columns and over 17K rows. Lots of the rows have NULL in some of the columns.
I want to remove the NULL so that queries return a blank ('') rather than a NULL.
Can I run some update function that replaces all NULL with '' ?
Using SQL Server 2008R2 Management Studio.
UPDATE my_table
SET column_1 = REPLACE (column_1,NULL,'')
But that would take forever to do it to all 59 columns!
What's the trick, team?
Use the SQL standard COALESCE:
UPDATE my_table
SET column_1 = COALESCE(column_1,'')
, column_2 = COALESCE(column_2,'')
...
WHERE column_1 IS NULL OR
column_2 IS NULL OR
... -- avoid empty updates
;
Then use ALTER TABLE ... to add NOT NULL constraints to all columns that shall not have NULL to prohibit re-introducing NULL values.
Don't use ISNULL, which basically is a duplication of the standard COALESCE in some RDBMS - and not available in others. (Well, there are subtle differences, read the manual for details or even more detail here.)
Of course, the empty string ('') is only valid for string types. Not for number types, for instance.
Use isnull function. Returns specified value if null is found in column
Select isnull(col1,'')
Try this one:
UPDATE yourtable SET column1 = ' ' WHERE column1 = NULL
Related
Here is what I am trying to do:
select * from table
where in ('completedDate', 'completedBy', 'cancelDate', 'cancelBy') is not null
If the four columns above are not null I need to display the records. I know I would do this with several where/and clauses but trying to learn and make my new stuff cleaner.
Is what I am trying to do above possible in a cleaner way?
If I understand correctly I guess you want to do that:
select *
from table
where completedDate is not null
and completedBy is not null
and cancelDate is not null
and cancelBy is not null
Regarding clarity of code I don't see a better way to write it, that's what I would code anyway.
EDIT: I wouldn't really do that in this case, but if this is a very common condition you can add a computed column in the table (stored or not), or create a view on top of table, and do:
select * from view where importantFieldsAreNotNull = 1
If I understand correctly, you want to return records where all four columns are not null?
The standard and (in my opinion) most readable way to do this would be:
Select
*
From
YourTable
Where
Column1 IS NOT NULL AND
Column2 IS NOT NULL AND
Column3 IS NOT NULL AND
Column4 IS NOT NULL;
To Check if all Columns are not null:
select * from table
where completedDate is not null
and completedBy is not null
and cancelDate is not null
and cancelBy is not null
You could use the COALESCE function to determine if all the column values were NULL.
The COALESCE function takes between 1 or more arguments and returns the first non-null argument. If at least one of the arguments passed into COALESCE is NOT NULL, then it will return that value, otherwise if all the arguments are NULL it returns NULL.
SELECT *
FROM TABLE
WHERE COALESCE(Column1, Column2, Column3, Column4) IS NOT NULL
Also depending on the datatypes of the columns, you may have to CAST them to the same datatype. For example, I wasn't able to use the COALECSE function on a DateTime column and a CHAR column without casting.
However, even though this would be shorter, I would not consider it "cleaner". I'd think it would be harder to read and maintain compared to having multiple ANDs in the WHERE clause.
-- Under reasonable assumption on data types:
select *
from [table]
where completedBy+cancelBy+DATENAME(yy,completedDate)+ DATENAME(yy,cancelDate)
is not null
I understand that AS is used to create an alias. Therefore, it makes sense to have one long name aliased as a shorter one. However, I am seeing a SQL query NULL as ColumnName
What does this imply?
SELECT *, NULL as aColumn
Aliasing can be used in a number of ways, not just to shorten a long column name.
In this case, your example means you're returning a column that always contains NULL, and it's alias/column name is aColumn.
Aliasing can also be used when you're using computed values, such as Column1 + Column2 AS Column3.
When unioning or joining datasets using a 'Null AS [ColumnA] is a quick way to make sure create a complete dataset that can then be updated later and a new column does not need to be created in any of the source tables.
In the statement result we have a column that has all NULL values. We can refer to that column using alias.
In your case the query selects all records from table, and each result record has additional column containing only NULL values. If we want to refer to this result set and to additional column in other place in the future, we should use alias.
It means that "aColumn" has only Null values. This column could be updated with actual values later but it's an empty one when selected.
---I'm not sure if you know about SSIS, but this mechanism is useful with SSIS to add variable value to the "empty" column.
When using SELECT you can pass a value to the column directly.
So something like :
SELECT ID, Name, 'None' AS Hobbies, 0 AS NumberOfPets, NULL AS Picture, '' AS Adress
Is valid.
It can be used to format nicely a query output when using UNION/UNION ALL.
Query result can have a new column that has all NULL values. In SQL Server we can do it like this
SELECT *, CAST(NULL AS <data-type>) AS as aColumn
e.g.
SELECT *, CAST(NULL AS BIGINT) AS as aColumn
How about without using the the as
SELECT ID
, Name
, 'None' AS Hobbies
, 0 AS NumberOfPets
, NULL Picture
Usually adding NULL as [Column] name at the end of a select all is used when inserting into another table a calculated column based on the table you have just selected.
UPDATE #TempTable SET aColumn = Column1 + Column2 WHERE ...
Then exporting or saving the results to another table.
This might be the situation in other databases as well but when you make the following query
SELECT * FROM MyTbl WHERE MyColumn != 'Foo'
then any record where MyColumn is, say, 'Bar' is fetched but not where MyColumn is NULL. I assume this is expected behavior and that there is a reason behind it and I'd like to know why.
Is NULL considered to be equal to 'Foo' or is it just not expected to be part of the condition because the condition (NULL != 'Foo') seems to be true.
In DB logic, NULL means that there is simply no defined data in this field. It's considered neither equal or different to anything. You have to filter on it explicitly if you want to fetch the relevant lines :
SELECT * FROM MyTbl WHERE MyColumn != 'Foo' OR MyColumn IS NULL
See Wikipedia.
In SQL Server works three-state logic, which mean that NULL = NULL UNKNOWN which treats as FALSE
In all relational databases, the null value is not equal to anything, including itself.
You would not be able to find the rows with null values for MyColumn even if your query was "select * from MyTbl where MyColumn = null". The only way to get them would be "select * from MyTble where MyColumn is null"
For a detailed explanation, see http://en.wikipedia.org/wiki/Null_(SQL)
Yep, as Tim commented, NULL values are unknown. SQL Server isn't going to make an assumption about a NULL value in order to make a match. You can check out more info about how SQL Server processes NULLs.
imagine there are 50 columns. I dont wan't any row that includes a null value. Are there any tricky way?
SQL 2005 server
Sorry, not really. All 50 columns have to be checked in one form or another.
Column1 IS NOT NULL AND ... AND Column50 IS NOT NULL
Of course, under these conditions why not disallow NULLs in the first place by having NOT NULL in the table definition
If it's SQL Server 2005+ you can do something like:
SELECT fields
FROM MyTable
WHERE stuff
EXCEPT -- This excludes the below results
SELECT fields
FROM MyTable
WHERE (Col1 + Col2 + Col3....) IS NULL
Adding a null to a value results in a null, so the sum of all your columns will be NULL.
This may need to change based on your data types, but adding NULL to either a char/varchar or a number will result in another NULL.
If you are looking at the values not being null, you can do this in the select statement.
SELECT ISNULL(firstname,''), ISNULL(lastname,'') FROM TABLE WHERE SOMETHING=1
This will replace nulls with string blanks. If you want another value use: ISNULL(firstname,'empty') for example. You can use anything where the word empty is.
I prefer this query
select *
from table
where column1>''
and column2>''
and (column3>'' or column3<'')
Allows sql server to use an index seek if the proper index/es exist. you would have to do the syntext for column 3 for any numeric values that could be negative.
I have a stored procedure with a parameter name which I want to use in a where clause to match the value of a column i.e. something like
where col1 = name
Now of course this fails to match null to null because of the way null works. Do I need to do
where ((name is null and col1 is null) or col1 = name)
in situations like this or is there a more concise way of doing it?
You can use decode function in the following fashion:
where decode(col1, name, 0) is not null
Cite from SQL reference:
In a DECODE function, Oracle considers
two nulls to be equivalent.
I think your own suggestion is the best way to do it.
What you have done is correct. There is a more concise way, but it isn't really better:
where nvl(col1,'xx') = nvl(name,'xx')
The trouble is, you have to make sure that the value you use for nulls ('xx' is my example) couldn't actually be a real value in the data.
If col1 is indexed, it would be best (performance-wise) to split the query in two:
SELECT *
FROM mytable
WHERE col1 = name
UNION ALL
SELECT *
FROM mytable
WHERE name IS NULL AND col1 IS NULL
This way, Oracle can optimize both queries independently, so the first or second part won't be actually executed depending on the name passed being NULL or not.
Oracle, though, does not index NULL values of fields, so searching for a NULL value will always result in a full table scan.
If your table is large, holds few NULL values and you search for them frequently, you can create a function-based index:
CREATE INDEX ix_mytable_col1__null ON mytable (CASE WHEN col1 IS NULL THEN 1 END)
and use it in a query:
SELECT *
FROM mytable
WHERE col1 = name
UNION ALL
SELECT *
FROM mytable
WHERE CASE WHEN col1 IS NULL THEN 1 END = CASE WHEN name IS NULL THEN 1 END
Keep it the way you have it. It's more intuitive, less buggy, works in any database, and is faster. The concise way is not always the best. See (PLSQL) What is the simplest expression to test for a changed value in an Oracle on-update trigger?
SELECT * FROM table
WHERE paramater IS NULL OR column = parameter;