DSXchange

Hi All

This is my problem.

Source is Decimal(18,6). Target is Decimal(7,3).

So if input is 000000002323.423567

result should be 2323.424.

Database: Oracle 8.0

DS version: PX 6.0

I am reading input as Varchar and using field() to extract values. I dont know how far i will be successful this way. This does not seem to be a good way of doing this.

Is there any other way to do this stuff.

Thanks in advance
Jay

Hi,

I found something for you.
Try this : DecimalToDecimal(StringToDecimal(convert(".", "", Link.FIELD))/1000, 'round_inf')/1000

000000000002.423567 -> 000000000002.424
000000000001.423467 -> 000000000002.423

Ciao
Matthieu

Just a precision.
I don't know why (may be something with the buffer) but in the transformer you use this formula you have to declare the field as decimal(11,3). If you really need to change to decimal(7,3) (not needed for Oracle) add a second transformer.

Bye
Matthieu

Thanks Matthew. I'll try it and let you know how it goes.

Hi Matthieu,

What is 'round_inf' doing ?

Thanks
Jay

Hi Matthieu

The conversion from Decimal(18,6) to decimal(7,3) is working when the input number=1234.123

But gives a 0000.000 when input number=12345.123456

I try to round(number,3) at source. If i view source, it shows as

decimal(38,10), i.e. 0000000000000000000000012345.1234560000

Any idea as to why its happening ?

PX 6.0 is the version of Datastage. Database is Oracle 8.0.

Thanks
Jay

Matthieu,

that is exactly the problem i am facing....

how do i fit in a decimal(8,3) in a column declared decimal(7,3) ?

to circumvent this problem,

i read the input as Varchar(8) to_char(round(input,3))

Now the problem is how to convert this string value to decimal ?

In another thread Ray mentioned some scenarios...i will let u know on Monday...

bye
jay

May I know the format of using round_inf ?

The format? Not knowing exactly what you need let's go with...

As noted earlier, it's one of several optional "rounding types" you can specify during functions that support rounding. The Type Conversion functions are here, scroll down to DecimaltoDecimal for the example discussed in this thread. And from the documentation: