[Scilab-users] arrhenius plot with two x-axis

Samuel Gougeon sgougeon at free.fr
Fri May 15 20:56:09 CEST 2015


> Hello,
> Le 15/05/2015 11:59, Krzysztof Lewandowski a écrit :
>>
>> I'm trying to make arrhenius plot of my data ('ln(S)' vs. '1000/T') 
>> with another x axis with 'T' on the top of the graph, but I have no 
>> idea how to do that. Is it possible to make such plot in scilab?
>>
> If the additional axis is linear, you may use drawaxis(): 
> http://help.scilab.org/docs/5.5.2/en_US/drawaxis.html
> Otherwise, newaxes() could be necessary: 
> http://help.scilab.org/docs/5.5.2/en_US/newaxes.html
Even with newaxes, it could be very difficult to match both bottom and 
top scales.
To turn this feasible and easy was somewhat the purpose of this report:
http://bugzilla.scilab.org/show_bug.cgi?id=7739
Its (too) implicit aim was to enable plots such that the bottom axis  is 
in wavelengths L and the top axis is in corresponding wavenumbers n=1/L 
or frequencies f=c/L
This kind of plot is very common in physics, but hard and tricky to get 
with scilab, at least when keeping the autotick mode.

After the report 7739, i think that the best way to enable such plots 
could be to add a callback attribute to the drawaxis object, that would 
point to a function F(x) computing its x'(x) values, with x spanning x 
data_bounds, and x'(x) tick being drawn at the x position in data scale. 
Depending on F(x), a smart x' grid could be quite difficult to draw, in 
particular for minor ticks (if any).

Samuel




More information about the users mailing list