Modelling with Spreadsheets
From ASCEND
Spreadsheets are an excellent tool for many quick-and-easy engineering calculations and even quite a lot of more complex calculations.
There are arguments against the use of spreadsheets though:
- they don't support units of measurement!
- you can't check your model 'at a view' because formulae are hidden and often written in terms of cryptic cell references such as 'H18' and 'AB23'
- they are saved in a binary format that makes reviewing changes difficult
- it is hard to reuse one small bit of an old spreadsheet that you previously made (without copying it or modifying it)
- it is hard to build modular structures
- it is hard to examine the nature of convergence problems in the Solver component (particular to Excel)
- it is hard to build automated testing into a spreadsheet[1]
See also the article "Spreadsheets Considered Harmful" at http://advosys.ca/viewpoints/2006/09/spreadsheet-risks/
[edit] References
- ↑ Test Reuse in the Spreadsheet Paradigm ftp://ftp.cs.orst.edu/pub/burnett/issre02.testReuseSpreadsheet.pdf

