How can I show decimals after division update

[joel]

I have a table which I want to update by dividing one field into
another. The update runs with no errors, but the results come out as
only a positive integer number.

The datatype for the result field is float.
The datatype for the other fields are int.

I have tried testing by dividing 7 by 10000, which should give me
0.0007. The result in the database is 0.

How do I define a field to show the full value with decimal positions?

If I enter .0007 into the field through the Enterprise Manager, it shows
0007. When I update by dividing, it shows 0.

Any help would be welcome.

Joel
*** Sent via Developersdex http://www.developersdex.com ***
Don't just participate in USENET...get rewarded for it!

[joel]

Hugo,

I started to list the columns in the table with their datatypes and as I
was doing this I realized that the datatypes for the fields I am
dividing are "int".

I changed them to decimal and it solved the problem.

If I hadn't gone through the exercise of listing the columns and their
datatypes I might have taken a lot longer to realize the problem.

I never had this problem with Access, which is why I did not recognize
the solution.

Thanks,

Joel

*** Sent via Developersdex http://www.developersdex.com ***
Don't just participate in USENET...get rewarded for it!

[Hugo Kornelis]

On 22 Apr 2004 13:57:12 GMT, joel wrote:

Hugo,

I started to list the columns in the table with their datatypes and as I
was doing this I realized that the datatypes for the fields I am
dividing are "int".

I changed them to decimal and it solved the problem.

If I hadn't gone through the exercise of listing the columns and their
datatypes I might have taken a lot longer to realize the problem.

I never had this problem with Access, which is why I did not recognize
the solution.

Thanks,

Joel

Hi Joel,

You don't need to change the datatype just to get the division
correct. If the columns themselves contain integer data, change them
back to int.

The division will also work correctly if you write

UPDATE ....
SET FloatCol = CAST(IntCol1 AS float) / IntCol2
WHERE ....

SQL Server will switch to non-integer division as soon as one of the
operands in of a float or decimal type. The cast explicitly converts
the IntCol1 to float before the division is carried out. Without the
cast, the division would have been made first (as integer division,
thus losing all fractions) and converted to float afterwards (to store
the result in FloatCol).

Best, Hugo
--

(Remove _NO_ and _SPAM_ to get my e-mail address)

[Josh Phillips]

Hugo,

I have been reading this thread and I am running into a similar problem.
I am using the CAST but I am still getting a return of 0.0 any help
would be greatly appreciated. Here is my code:

select cast(count(Safe tyTrainingYN) AS FLOAT)
from ICPCDATA
where SafetyTrainingY N = 1 /
(select count(SafetyTra iningYN)YES
from ICPCDATA)

remove _nospam_ for my email

*** Sent via Developersdex http://www.developersdex.com ***
Don't just participate in USENET...get rewarded for it!

[Hugo Kornelis]

On 29 Apr 2004 17:57:18 GMT, Josh Phillips wrote:

Hugo,

I have been reading this thread and I am running into a similar problem.
I am using the CAST but I am still getting a return of 0.0 any help
would be greatly appreciated. Here is my code:

select cast(count(Safe tyTrainingYN) AS FLOAT)
from ICPCDATA
where SafetyTrainingY N = 1 /
(select count(SafetyTra iningYN)YES
from ICPCDATA)

remove _nospam_ for my email

*** Sent via Developersdex http://www.developersdex.com ***
Don't just participate in USENET...get rewarded for it!

Hi Josh,

I can't test right now, but it looks like a parethesis problem.

(I assume the extra "YES" is a typo? Otherwise, this should give you a
syntax error!)

The server will count the number of non-NULL values in
SafetyTrainingY N in ICPCDATA. It will then calculate 1 / that number,
as integer division (yielding 0). Next, it will find rows in ICPCDATA
with SAfetyTrainingY N = that result (0). Apparently, there are none,
so the count(..) yields 0. This is converted to float and the result
is given to you: 0.0

What you probably meant is:

select (select cast(count(Safe tyTrainingYN) AS FLOAT)
from ICPCDATA
where SafetyTrainingY N = 1) /
(select count(SafetyTra iningYN)
from ICPCDATA)

By the way, why don't you use 'Y' and 'N' instead of 1 and (I think) 2
to represent yes and no? You even have Y and N in the columns' name!

Best, Hugo
--

(Remove _NO_ and _SPAM_ to get my e-mail address)

[Josh Phillips]

Hugo,

I have been reading this thread and I am running into a similar problem.
I am using the CAST but I am still getting a return of 0.0 any help
would be greatly appreciated. Here is my code:

select cast(count(Safe tyTrainingYN) AS FLOAT)
from ICPCDATA
where SafetyTrainingY N = 1 /
(select count(SafetyTra iningYN)YES
from ICPCDATA)

remove _nospam_ for my email

*** Sent via Developersdex http://www.developersdex.com ***
Don't just participate in USENET...get rewarded for it!

[Hugo Kornelis]

On 29 Apr 2004 17:57:18 GMT, Josh Phillips wrote:

Hugo,

I have been reading this thread and I am running into a similar problem.
I am using the CAST but I am still getting a return of 0.0 any help
would be greatly appreciated. Here is my code:

select cast(count(Safe tyTrainingYN) AS FLOAT)
from ICPCDATA
where SafetyTrainingY N = 1 /
(select count(SafetyTra iningYN)YES
from ICPCDATA)

remove _nospam_ for my email

*** Sent via Developersdex http://www.developersdex.com ***
Don't just participate in USENET...get rewarded for it!

Hi Josh,

I can't test right now, but it looks like a parethesis problem.

(I assume the extra "YES" is a typo? Otherwise, this should give you a
syntax error!)

The server will count the number of non-NULL values in
SafetyTrainingY N in ICPCDATA. It will then calculate 1 / that number,
as integer division (yielding 0). Next, it will find rows in ICPCDATA
with SAfetyTrainingY N = that result (0). Apparently, there are none,
so the count(..) yields 0. This is converted to float and the result
is given to you: 0.0

What you probably meant is:

select (select cast(count(Safe tyTrainingYN) AS FLOAT)
from ICPCDATA
where SafetyTrainingY N = 1) /
(select count(SafetyTra iningYN)
from ICPCDATA)

By the way, why don't you use 'Y' and 'N' instead of 1 and (I think) 2
to represent yes and no? You even have Y and N in the columns' name!

Best, Hugo
--

(Remove _NO_ and _SPAM_ to get my e-mail address)