A Note on Filenames & Directories

Filenames

CIAO imposes very few restrictions on valid filenames. Beyond any restrictions imposed by the operating system, generally CIAO file names cannot contain square braces, [], as they are used as part of the CIAO filtering syntax. Additionaly file names or directories with spaces are also generally problematic as they may get stack expanded into invalid parts. Often the special value 'none' is used as a special value when a file is optional. Finally, users should avoid file names with a plus sign, +, especially if followed exclusively by numbers, for example "my_cluster+32"; which may cause the underlying CFITSIO library to attempt to open the trail numbered extension of the file, and likely fail.

The file names given in the threads are those for the test data used in the examples. The unix mv or ln commands may be used to change file names as desired. For instance, one could link them into a working directory with the names pha, rmf, asol, evt1, and bpix. The same script could then be run several times without having to input the observation-specific filenames.

For further information on CXC file naming conventions, please see the FITS File Names for the Chandra Data Archive and the Guide to Chandra Data Products pages.

Directories

Similarly, it is not required that you keep the files organized in the "primary" and "secondary" directories in which they are distributed. The data products are arranged such that the products necessary for most analyses (e.g. the CIAO threads) are in the primary directory. The Data Products Guide contains more details on the data distribution.

Some files contain references to filenames of auxiallary files needed for futher analysis. These include things such as the PHA (spectrum) file that contains references to the ARF, RMF, and background; as well as the event file which contains references to a several files need for later analysis. Tool may fail if these file names are changed and the headers of the files not updated.