That depends largely on what you are trying to do with the data, what format you are acquiring the data in and how much coding experience you have. 

   Python will be the easiest to start with and has a huge level of scaleability which if you are using excel macros now will more than adequately cover your needs. 

   Are you running statistics for publication or are you trying to integrate the data with additional software or hardware?

Jon

MySQL is the excellent DBMS tool, Platform independent and suits to all major programming languages.

If you don't know coding very well then you may use navicat MySQL GUI software 

Programming Languages are just like path to a destination. Once we reach it is the application and the supporting OS which do the job.

For routine calculations and probably from large database retrieval, almost all languages have same efficiency ( unless you are looking for real time , high performance solutions ). 

Just the RDBMS and its interface ( drivers ) for the language matters. You may try java as it has database functionality for most of the DBMS. For a standalone ( desktop ) application .Net/VC++/VB/Java do the same job. For Web interface choose python/js.

Choose MySQL for generic RDBMS and  Postgres SQL with POSTGIS if you are looking for spatio temporal analysis. I understand that solar radiation at different geographic locations ( lat / lon ) plays a major role in your analysis. Go for Postgres option right from the beginning.

If part of the problem with using Excel is the long computation time, Python is probably not the best solution, since it tends to be rather slow itself. Both C and Fortran can be used for fast computations, but they take some time to learn. Especially, because you will not only need not write a program, but to optimize for high computation speed as well.

Do you know anyone  who can help with the programming? If not, I would think hard before starting such a major project. You may lose as much time learning how to use the programming language as you gain from the higher computation speed. On the other hand,  programming is a very useful and transferable skill, so if you expect to use it again for other purposes, it would certainly be worth the time and effort

I recommend PowerBasic. You can transfer your VBA code very easy to PB. This is extremely fast language, like C++. It has powerful graphic also and offers parallel programming. 

http://en.wikipedia.org/wiki/PowerBASIC

I can offer you such option. I can run some of your macros on my computer under PowerBasic so you can see how the calculation speed increases under PB.

IDL

I think that you should consider either Python or MatLab. Those are fitted with several Maths functions already implemented which if you consider future scaleability it will save you alot of time, afterwards. Both are easily wrapped/linked with other programming languages such as C and Java. Python it's free, MatLab is licensed but fits other funcionalities, then you choose ;)

I have carefully read all of the answers that have come... all very interesting. I am trying to understand the answer of Dr. Breuer.

 Your responses have given me a perspective much more extensive that I had so far. Thank you very much. I always (7 years)  processed my data with spreadsheet (Excel) and I think that is time to learn new techniques, but the philosophies of some softwares are so different among them so I will start with something familiar (Origin)  and see what's next (perhaps Matlab)

A very strong calculating tool-language is R. 

Probably you will find some easy to use packages of R scripts.

Physical systems are best modelled e.g. with MapleSim, or Wolfram's System Modeller.

We use an Excel add-in that is available on our website.  It is programed in basic, which is not very practical any more.  If you want others to be able to use the program and have it operate on different platforms, C would probably be the best language.  If you are using very large datasets, then a database is called for and the programing should comply with the database requirements.

Have you ever considered bigdata implementation?

In addition  you can follow this link to see a ready to use package for solar photovoltaic systems calculations in R language.

http://cran.r-project.org/web/packages/solaR/solaR.pdf

If I were in the same situation I would use either Matlab or R because of their strong support base and the quantity of freely available routines for solar applications. See e.g.

https://pvpmc.sandia.gov/resources-for-members/pv_lib-toolbox/

http://procomun.wordpress.com/

http://oscarperpinan.github.io/solar/

http://www.r-bloggers.com/maps-of-solar-radiation/

To decide between those two (or any other solution previously mentioned), I would ask around and try to find a colleague who can help smooth the learning curve.

Excel should be your last choice. Given that your data analysis requires now a more advanced language for you to learn, I strongly recommend learning Python rather than Origin or Matlab. Python is open source and is becoming increasingly popular amongst scientists for its simple syntax (as compared with other languages) and its impressive set of scientific oriented libraries (like numpy and scipy). For an optimized workflow, I recommend you to start with IPython (which uses the web browser to edit and run Python code).

It would be most helpful if you could describe in more detail

I also agreed with some answers above. R programming language is free and the simple calculation, as you mentioned, is easy to run with some basic clauses even if no complicated packages are installed, such as ddply.

Personally, we have been using MATLAB to process satellite data records of surface vegetation. MATLAB is great with GUI and libraries. The only downside is that it is a propriety software and hence, not free.

However, MATLAB do have performance issues. Particularly, that involving a large amount of looping as could be seen in some optimization algorithms. Hence, we are now partly moving our algorithms to parallel processing platforms.

I know people use C/C++ for operational/production systems due for performance but that may not be ideal during algorithm development.

hola, depende cual sea el origen de tus datos, si provienen de imagenes satelitales en la universidad programamos en basic, cuando son datos de estacion o track usamos matlab..

I am currently working with the "Tracer", an open source Python-based ray-tracing program. The program is originally from Yosef Meller and I continue its development while  I develop my PhD work.

URL: https://github.com/casselineau/Tracer

It uses Numpy and consequently runs faster than could be expected with Python only. The big advantage is that you can do what you want with your ray-tracing because the program is completely open. I personnally use it coupled with hydrodynamic models to optimise concentrated solar receiver designs, something that is quite uneasy to do with commercial softwares for example.

https://github.com/casselineau/Tracer

https://www.researchgate.net/profile/Yosef_Meller

Dear sir,

with best wishes, please keep the data on excel and use Matlab to do the calculations you can add your calculations to excel file you already have.

best of luck,

saeid parsa

Interactive Data Language (IDL) works great for programming large data sets.  Not only can you use the built-in statistical packages, but this language handles a large, matrix of information with ease.

Modelling and simulation of solar radiation data processing with ...

https://www.semanticscholar.org/...solar-radiation-data.../dbb6d76b3a68ced86a30d413f...

Thank you, Abhijit. Unfortunately, I can't enter the address you sent !

I strongly recommend the python "pvlib" package.

https://pvlib-python.readthedocs.io/en/stable/index.html

Thank you very much, Inti. We use Python a lot.

But also we use Fortran, Matlab and R.

Python

Thank you Brighton !

Python is becoming the programming standard for solar radiation data analysis.

SolaR in R is a good alternative too.

Thank you Ian

Good discussion