Updated readme, install files, minor code cleanups

[webbpinner]

I saw this project was being worked on again and took a look to see how things have changed. I forked the project again and tried to add some changes to improve use by other groups.

• I noticed that the preferred file format has moved from csv to pickle. I think this was a great decision but I know some folks will still want to produce a csv-formatted version. I've added this as an option to odf_convert_sbe and odf_process_bottle. The default behavior is still to create a pickle-formatted file, just added the option to also produce a csv-formatted version

• I noticed an issue in the in odf-process-bottle where the dataType for all columns was set to float64 as a result of calculated a mean. I added some code to correctly re-typecast certain columns as bool or int as appropriate.

• Added install instructions for Mac OS X to INSTALL.md

• Updated README.md

• Updated CTD_PROCESS.md to use markdown formatting. This will make the file more easily read on GitHub

• Incorporated ODF-CTD-PROC dependencies in INSTALL.md

• Restored sample_convert, sample_import and sample_export scripts for the purposes of testing ODF_CTD_PROC scripts and libraries.

• It would be very useful if ODF could add some of their recent CTD data to the repo for testing purposes. It would also be greatly appreciate it if ODF could upload a configuration.ini file... again for testing/development purposes.

• General Code cleanup.

I hope these changes were seen as helpful to the project.

[asx-]

Hi Webb,

We're currently in the process of big architectural change right now, both in code and in organizationally. I can't accept any PRs right now since there's a lot of discussion about how we want to shape and restructure this project for the future. Once some things settle down organizationally and we can get some basic standards documents in place, we'll be open for PRs again.

Quick comments on the code: Although it's not in place yet, we want to get away from any print/debugPrint statements and move completely to the logging module. CSV primarily served as the interchange format first, data inspection second. That format will most likely change away from pickle towards netCDF via xarray when we're not making urgent patches to the code to keep things running, which requires having a routine to read netCDF, but will be much easier to support than CSV.

[webbpinner]

Thanks for the reply and explanation. I do agree this size of this project now warrants a redesign.

I'm encountering more and more users that have a need for a library/tool that can enable them to develop semi-automated processing of SBE data files.

Please let me know if there's anything I can do to help.

[asx-]

Closing this for now because I think enough change has happened to package structure, files to require a serious update that would be better dealt with after the architecture settles down more. Will revisit in January 2018