Chronology | Current Month | Current Thread | Current Date |
[Year List] [Month List (current year)] | [Date Index] [Thread Index] | [Thread Prev] [Thread Next] | [Date Prev] [Date Next] |
On 12/10/2011 09:58 AM, I wrote:
science
/ \
/ \
theory experiment
/ \
/ \
algebra simulation
The question arises, what about visualization?
The answer is that visualization is important. It's importance
spans all the categories given above:
-- You can visualize an algebraic formula. For example, it may be
worth plotting 0.5 a t^2 versus t, as a model of the motion of a
freely-falling object. As a slightly more advanced example,
consider the Shockley diode equation. It really pays to plot
that, because the concept of "diode drop" is not built into the
equation, but emerges as a none-too-obvious consequence.
-- You can visualize simulation results.
-- You can visualize experimental data.
This is relevant to the current thread, at least tangentially, because
often computers provide the best way of visualizing some result ...
including /but not limited to/ simulation results.
In particular, plotting stuff with ye olde spreadsheet application is
a good example of doing science with a computer. It has the advantage
of being interactive: If you change the data, the graph changes immediately.
It also has the advantage of not having much barrier to entry; it does
not require any fancy programming language.
On 12/10/2011 06:00 PM, Forinash III, Kyle wrote in part:
Simulations have to be interactive. They can't be just visualizations.
I agree with the sentiment, but I might have worded it differently.
a) visualizing an interactive model is good.
b) visualizing a one-shot static model is not so good.
The problem with (b) is not the visualization; the problem is the
inflexible model.
It's not worth arguing about whether it is better to have a good
model without good visualization or vice versa; I reckon you need
*both* a good model and good visualization; otherwise the whole
enterprise is not worth the trouble.
===========================