We need to use a parsing a Math Expression Parser int two parts of our
system. Doing that, we will need to use the two "Flavors" of the Jep
librairy (Java and .Net).
Does the Jep java and the Jep.Net support the same set of
functionnalities with the same result or there's some limitation?
Also, I saw in the documentation that the Jep java have some issues
regarding the accurency of some calculation (ie: 10*0.09 evaluates as
0.8999999999999999) unless you use the new limitating BigDecimal mode.
Does the Jep.Net have the same kind of issue ?
I'm not particularly up to speed with Jep.Net but it does follow the
more recent Jep version 3 branch but is a couple of minor revisions
behind, so probably 3.0 rather than 3.4. So you should get the
majority of function of the java version. If there are particular
things you need we could probably provide those.
As to the limits of accuracy, the Jep behaviour is exactly the same as
using the double primitive data type. This holds in both Java and C#.
For scientific and engineering applications this is the most
appropriate data-type to use. Appropriate formatting, using printf or
similar will display the numbers correctly. For financial calculations
you will probably want the BigDecimal type, most of the limits of this
type (such as not being able to use the sin function) don't affect
financial calculations. I'm not 100% sure how BigDecimal is
implemented in the C# version, I suspect it might use the C# decimal
type.
Hope thats of help
Richard
On Dec 22, 3:34 pm, Boul <jf.b...@gmail.com> wrote:
> We need to use a parsing a Math Expression Parser int two parts of our
> system. Doing that, we will need to use the two "Flavors" of the Jep
> librairy (Java and .Net).
> Does the Jep java and the Jep.Net support the same set of
> functionnalities with the same result or there's some limitation?
> Also, I saw in the documentation that the Jep java have some issues
> regarding the accurency of some calculation (ie: 10*0.09 evaluates as
> 0.8999999999999999) unless you use the new limitating BigDecimal mode.
> Does the Jep.Net have the same kind of issue ?
Modification to previous post. Jep.Net does not have an equivalent to
BigDecimal or decimal type. It would not be too much work to modify
the code to add these features.
Richard
On Dec 22, 3:34 pm, Boul <jf.b...@gmail.com> wrote:
> We need to use a parsing a Math Expression Parser int two parts of our
> system. Doing that, we will need to use the two "Flavors" of the Jep
> librairy (Java and .Net).
> Does the Jep java and the Jep.Net support the same set of
> functionnalities with the same result or there's some limitation?
> Also, I saw in the documentation that the Jep java have some issues
> regarding the accurency of some calculation (ie: 10*0.09 evaluates as
> 0.8999999999999999) unless you use the new limitating BigDecimal mode.
> Does the Jep.Net have the same kind of issue ?
I played a little with the .net version and I find out that the
StringFunctionSet component (http://www.singularsys.com/jep/doc/html/ functions.html#string) seems to be missing in the .net version.
Am I right?
Jeff
On Dec 22 2011, 3:47 pm, Richard <sings...@gmail.com> wrote:
> Modification to previous post. Jep.Netdoes not have an equivalent to
> BigDecimal or decimal type. It would not be too much work to modify
> the code to add these features.
> Richard
> On Dec 22, 3:34 pm, Boul <jf.b...@gmail.com> wrote:
> > Hi,
> > We need to use a parsing a Math Expression Parser int two parts of our
> > system. Doing that, we will need to use the two "Flavors" of the Jep
> > librairy (Java and .Net).
> > Does the Jep java and the Jep.Netsupport the same set of
> > functionnalities with the same result or there's some limitation?
> > Also, I saw in the documentation that the Jep java have some issues
> > regarding the accurency of some calculation (ie: 10*0.09 evaluates as
> > 0.8999999999999999) unless you use the new limitating BigDecimal mode.
> > Does the Jep.Nethave the same kind of issue ?
