I am looking for a way to create average curves for plotting rheological data in a log/log graph, similar to the way in which spectroscopy software averages multiple spectra (e.g. from mapping) into a single curve. Excel seems limited in this regard, so any tips regarding alternative software (for Mac and preferably free or widely available) would be appreciated.
Have you tried Origin? It can do it. It's not free but worth having...
http://www.originlab.com/
I would also recommend Origin for plotting and analysing data as it has a lot of built in functions that are intuitive to use
Thank you both for the recommendation. I know of Origin but haven't tried it yet. Please could you let me know the name of the function that enables this so that I can check that it is suitable before trying it?
I'm currently trialling GraphPad Prism. It's possible with this to enter multiple y values for experimental replicates, but requires the x values to be constant between replicates. This is fine for something like FTIR, where there's a reading for each fixed wavenumber, but for mechanical tests such as the rheological data I'm trying to process into a single curve, the x values vary as well as the y values. Do you know if this is also a limitation in Origin?
In Prism, I've also tried using the Lowess function to average the curves, but even the "fine" setting returns crude graphs with a large shift to the right on the x axis.
NIck, this might work: http://www.originlab.com/doc/Origin-Help/Math-AveCurve
There are a number of ways for doing it. Perhaps, the easiest would be to import data into the Origin spreadsheet and calculate the new column as the average of multiple input columns using the function "Set column values..." from a drop down menu of "Columns". You can also highlight the input columns and see "Statistics on Columns" which gives various statistical data including the average. Also, you can plot all the individual input data on a single graph and select "Analysis/Mathematics/Average Multiple Curves" which will do the job.
Thank you both for the further advice. Bharath, the link you sent is useful.
Could either of you please confirm: does Origin deal with varying x values as well as varying y values?
Yes, it does. In principle, you can do such averaging even for data sets that have different x values. You'd need to designate some column as x and then interpolate y values from all data sets to the reference x values. Also, the increment both in x and y can be varying.
In addition to the link provided by Bharath, the following link may be useful as well:
http://www.originlab.com/forum/default.asp
Good luck!
If you want a powerful freeware solution, try R, the statistical programming language. You have every fitting function possible, but you do not have a point & click interface. There's no free lunch :-).
Thanks Alan. Sounds good but probably beyond my mathematical/computing skills!
Update: the average graph function in Origin works amazingly, thanks very much!
Dear Dr Walters:
I agree with users of the Origin, despite of another friendly softwares.
Here, I use for plots and simulations an old version of the Origin, since is very easy to entry with a simple or complex formula.
Please, if possible consider seeing the paper recently published by us in attach, in which Equations 11 and 12 were implemented in the Origin to attain further simulation/verifying of a physical response of a ceramic material. Results derived from the Origin (theoretical adjust) are showed in Fig. 13 of the paper.
In the sequence, a plot with multiple graphics can be observed in Fig 14. See, the scale presentation can be adjusted by a simple click linear x log, linear x ln, log x log etc,...,etc and turned to the original again by a simple click.
Kind regards
Marcos Nobre
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Graphs