I know that material studio can convert the PXRD data of material into crystal cell parameters and give idea about the structure of that material. But how to initialize those calculations?
Never use a software when you do not understand what and how it is doing something. In other words, read some literature or books to get an imagination, how these data are related to each other (XRD and crystallography) and then do not start to write your own software but look for some which is doing this you are interested in. Material Studio possibly can do it but each software is only that smart as its operator. You will certainly get a number, but you know that every software has bugs, and nobody knows where they are and when they become essential. If a software gives as result of 2*3=5, you can believe it (must be true since it is given by the software)... or you know it is wrong and you expect that there obviously the programmer mistyped * and +. No error message since everything is fine...
Gert Nolze is absolutely right. Crystallography is not an easy subject, but luckily enough there exist many computer programs that work well -- if you know that subject.
To solve a crystal structure from PXRD data from scratch is certainly not something that even experienced crystallographers gladly embrace, mostly because PXRD does not give you all information that you need, for instance not even the crystal system! What PXRD can give you (sometimes after considerable effort) is a suggested unit cell, but the true symmetry cannot be deduced.
Mathematical relations may be found that seem convincing, but there are devious pitfalls also! I once indexed a pattern (Sc3As2) with very high precision on a tetragonal cell, but later -- through comparison with other similar chemical systems -- I realized it had to be orthorhombic: Two of the axes were equal within 0.02%! There was no fourfold symmetry (Space group Pnma), and assuming that there were (because of the "successful" indexing) could never yield the correct structure.
It may be difficult enough to solve a structure from single-crystal data...
Another possibility is to use FlexCryst, which is been developed by us. It works quite simple and in automatic way. You might send me your molecule and I can estimate, if the software fits for you. Below I attached a picture from the graphical user interface.
@Nolze and Rolf Berger your suggestions are really true but not solve of Sandeep's Q. Must be people will use software after learning not before.......