G
Guest
I have a number of datasets that contains lots of engineering type data.
Things are in lengths, forces, energies, times, etc etc.
The datasets have two underlying natures:
1. The product dataset has all numeric columns and contains only values of
the lengths, etc. It does not contain the associated units because they are
only column dependent. (All values in a column have same kind of dimension of
course such as length but also the same actual dimension such as feet or
pounds etc.)
2. User input dataset is all string based. Each value here carries along its
associated dimension because some may be in feet and some in inches etc. We
want to carry along the exact user input as a string converting it into
numbers for the crunch but displaying the original input in reports.
SOOOO ... Short of a lot of hand coding - which this project specifically
wants to avoid -
A. In both cases is there a way of creating types say "energy1" and
"energy2" whose underlying representation is double in case 1 and string in
case 2 and using that as the datatype of a column. (I think this is doable
and maybe even trivial, but am not sure how.)
B. Much more difficult, may be impossible, is there any reasonable way to
add additional meta data to a column so that in case 1 I could not only type
the data as energy1 but also have the additional meta data say that the
numbers all have the units of ft / sec / sec (or whatever the units of energy
are. I forget at the moment.)
Basically is it possible to have the dataset and its metadata document the
underlying reality through types and metadata without the expense of have
extra columns in the table to contain the information. (One could imagine
having for each column of real data another one that contains the dimension
family and the actual dimension e.g. "Length:feet" This is what I want to
avoid.
--
Regards,
Al Christoph
Senior Consultant and Proprietor
Three Bears Software, LLC
just right software @ just right prices @3bears.biz
Things are in lengths, forces, energies, times, etc etc.
The datasets have two underlying natures:
1. The product dataset has all numeric columns and contains only values of
the lengths, etc. It does not contain the associated units because they are
only column dependent. (All values in a column have same kind of dimension of
course such as length but also the same actual dimension such as feet or
pounds etc.)
2. User input dataset is all string based. Each value here carries along its
associated dimension because some may be in feet and some in inches etc. We
want to carry along the exact user input as a string converting it into
numbers for the crunch but displaying the original input in reports.
SOOOO ... Short of a lot of hand coding - which this project specifically
wants to avoid -
A. In both cases is there a way of creating types say "energy1" and
"energy2" whose underlying representation is double in case 1 and string in
case 2 and using that as the datatype of a column. (I think this is doable
and maybe even trivial, but am not sure how.)
B. Much more difficult, may be impossible, is there any reasonable way to
add additional meta data to a column so that in case 1 I could not only type
the data as energy1 but also have the additional meta data say that the
numbers all have the units of ft / sec / sec (or whatever the units of energy
are. I forget at the moment.)
Basically is it possible to have the dataset and its metadata document the
underlying reality through types and metadata without the expense of have
extra columns in the table to contain the information. (One could imagine
having for each column of real data another one that contains the dimension
family and the actual dimension e.g. "Length:feet" This is what I want to
avoid.
--
Regards,
Al Christoph
Senior Consultant and Proprietor
Three Bears Software, LLC
just right software @ just right prices @3bears.biz