043c660 | Brandon Rhodes | 20 December 2016, 05:32:49 UTC | Fix test to match de405 we actually download In other news: I had forgotten that the version of de405 that this project's `.travis.yml` downloads from http://naif.jpl.nasa.gov is different from the one that Skyfield gets from the ssd.jpl.nasa.gov server. I kind of wish that there were only one version of DE405. | 20 December 2016, 05:32:49 UTC |
cc3bb7f | Brandon Rhodes | 20 December 2016, 05:22:36 UTC | Fix #21 by repairing broken command line; add test | 20 December 2016, 05:22:36 UTC |
7141d34 | Brandon Rhodes | 20 December 2016, 05:13:09 UTC | Include instructions for fixing a test skip | 20 December 2016, 05:13:09 UTC |
001709d | Brandon Rhodes | 08 June 2016, 18:14:05 UTC | Remove old conda recipe in favor of feedstock | 08 June 2016, 18:14:05 UTC |
40afdb4 | Brandon Rhodes | 08 June 2016, 05:17:04 UTC | Fix #18 by adding explicit close() method This relies on reference counting (or garbage collection) to destroy the memory map segments hiding behind the various NumPy arrays that are sitting behind each Segment. In the future we might want to be more proactive about getting each memory map destroyed immediately too. Happily, the close() calls this patch makes during tests also eliminate several "ResourceWarning: unclosed file" warnings that print out when the tests are run under Python 2.5. | 08 June 2016, 05:17:04 UTC |
6b2ed1e | Brandon Rhodes | 03 April 2016, 00:25:01 UTC | Switch setup.py `install_requires` to `requires` Will hopefully fix #16. | 03 April 2016, 00:25:01 UTC |
c66d674 | Brandon Rhodes | 03 April 2016, 00:22:31 UTC | Stop fetching deprecated ephemeris package There is really no point in having Travis burn bandwidth for this on every commit any more. | 03 April 2016, 00:22:31 UTC |
7891d59 | Brandon Rhodes | 03 April 2016, 00:22:01 UTC | Add Python 3.5 to Travis configuration | 03 April 2016, 00:22:01 UTC |
6e244b3 | Brandon Rhodes | 31 March 2016, 12:05:54 UTC | Update gitignore file | 31 March 2016, 12:05:54 UTC |
368abd7 | Brandon Rhodes | 31 March 2016, 12:04:04 UTC | Fix JPL report links in old ephemeris READMEs Apparently JPL report URLs sometimes change. I will not be uploading new versions of these ephemerides, and maybe should remove them someday from PyPI, but will run "setup.py register" to update the links for now. | 31 March 2016, 12:04:04 UTC |
7c76b16 | Brandon Rhodes | 09 November 2015, 21:53:37 UTC | Support getting summaries from StringIO files | 09 November 2015, 21:53:37 UTC |
f6ec306 | Brandon Rhodes | 02 November 2015, 02:20:41 UTC | Update documentation and bump version to 2.4 | 02 November 2015, 02:20:41 UTC |
1c02f0e | Brandon Rhodes | 02 November 2015, 02:16:35 UTC | Fix #13 by switching to "ALLOCATIONGRANULARITY" | 02 November 2015, 02:16:35 UTC |
13a308d | Brandon Rhodes | 19 October 2015, 13:36:47 UTC | Prevent deprecation warning for float in `shape` | 19 October 2015, 13:37:29 UTC |
f340ca5 | Brandon Rhodes | 19 October 2015, 13:35:06 UTC | Add names for timescales now included in DE430 | 19 October 2015, 13:35:06 UTC |
a8bd0df | Brandon Rhodes | 17 August 2015, 06:02:23 UTC | Bump version and documentation to version 2.3 | 17 August 2015, 06:02:23 UTC |
f00421d | Brandon Rhodes | 16 August 2015, 05:03:38 UTC | Roll legacy DAF support into main class | 16 August 2015, 05:03:38 UTC |
2b68f30 | Brandon Rhodes | 16 August 2015, 05:03:28 UTC | Fix the docstring for the SPK class | 16 August 2015, 05:03:28 UTC |
1f3db14 | Brandon Rhodes | 05 August 2015, 15:19:20 UTC | Final tweaks that bring it to version 2.2 | 05 August 2015, 15:19:20 UTC |
93f7e54 | Brandon Rhodes | 04 August 2015, 13:07:24 UTC | Add de405 to kernels downloaded for Travis CI test | 04 August 2015, 13:07:24 UTC |
3e80ee7 | Brandon Rhodes | 04 August 2015, 03:38:47 UTC | Add a function for reading an ASCII header file Fixes #9. | 04 August 2015, 03:38:47 UTC |
b5d3a5f | Brandon Rhodes | 04 August 2015, 03:22:09 UTC | Add support for old NASA kernels like de405.bsp Fixes #12. | 04 August 2015, 03:23:28 UTC |
33bf5a8 | Brandon Rhodes | 04 August 2015, 03:04:56 UTC | Simplify SPK constructor by demanding DAF argument | 04 August 2015, 03:23:28 UTC |
7884c80 | Brandon Rhodes | 04 August 2015, 03:18:25 UTC | Merge pull request #10 from Juanlu001/conda-recipe Added conda recipe | 04 August 2015, 03:18:25 UTC |
0f0bf19 | Brandon Rhodes | 23 June 2015, 03:12:35 UTC | Fix typo pointed out in #12 | 23 June 2015, 03:12:35 UTC |
f1045ac | Brandon Rhodes | 02 June 2015, 14:58:26 UTC | Fix exception caused by coefficient_count=2 | 02 June 2015, 14:58:26 UTC |
74d1175 | Brandon Rhodes | 02 June 2015, 14:25:54 UTC | Fix final test message for Python 2.6 | 02 June 2015, 14:25:54 UTC |
bb174e3 | Brandon Rhodes | 02 June 2015, 13:42:03 UTC | Expand jpltest.py to understand targets 11, 12, 13 | 02 June 2015, 13:42:03 UTC |
36c0cc0 | Brandon Rhodes | 02 June 2015, 03:47:59 UTC | Remove unused import | 02 June 2015, 03:47:59 UTC |
397ca0d | Brandon Rhodes | 20 April 2015, 16:09:43 UTC | Add changelog entry | 20 April 2015, 16:09:43 UTC |
7e310bd | Brandon Rhodes | 20 April 2015, 16:08:05 UTC | Add a simple command line to the module | 20 April 2015, 16:08:05 UTC |
7a7c3b9 | Juan Luis Cano RodrÃguez | 04 April 2015, 18:20:57 UTC | Added conda recipe | 04 April 2015, 18:20:57 UTC |
2e74682 | Brandon Rhodes | 24 February 2015, 06:55:51 UTC | Update changelog and version for 2.1 | 24 February 2015, 06:55:51 UTC |
26c75cd | Brandon Rhodes | 13 February 2015, 11:41:20 UTC | Issue friendlier error for files not in 'b' mode | 13 February 2015, 11:41:20 UTC |
1fd914f | Brandon Rhodes | 13 February 2015, 11:28:47 UTC | Switch from one big mmap() to an mmap() per array Bringing the whole file in at once was offering the error OSError: [Errno 12] Cannot allocate memory in experiments I was doing with Skyfield with Jovian moons, because of the large size of jup310.bsp plus some other data that I was using. So with this change, jplephem uses mmap() separately for each segment the user needs, making it possible to map only one — or a few — segments of a file at a time, discarding older ones as newer ones are needed and limiting the amount of simultaneously mapped memory required. | 13 February 2015, 11:28:47 UTC |
5a4e7ab | Brandon Rhodes | 08 February 2015, 20:20:04 UTC | Have Travis CI do a real install from .tar.gz | 08 February 2015, 20:20:04 UTC |
4f7a3aa | Brandon Rhodes | 08 February 2015, 20:17:22 UTC | Final documentation tweaks for 2.0 | 08 February 2015, 20:17:22 UTC |
32a51f5 | Brandon Rhodes | 08 February 2015, 19:31:07 UTC | Change main computation routine to a generator This is kind of magic. Because it is now a generator, you have the option of asking for position, then only proceeding to the velocity if you need it, without having to repeat the work of getting to that point again. | 08 February 2015, 19:31:07 UTC |
2f7dfef | Brandon Rhodes | 08 February 2015, 19:04:49 UTC | Further document the new API, in code and README | 08 February 2015, 19:04:49 UTC |
c4575cc | Brandon Rhodes | 08 February 2015, 18:03:19 UTC | Pivot to index segments by both center and target The great wonder of writing documentation once again asserts itself! Only while writing up the documentation — frighteningly enough, the last step before release — did I realize that the API was dangerously misleading because it let you just ask the position of a target, without any context as to where the measurement was from. So the quick go-to indexing that the SPK object supports — which I now like so very much that I have promoted it to the __getitem__ of the kernel object — now makes the user name both the center and the target that they want a vector between. | 08 February 2015, 18:03:19 UTC |
7b0ceef | Brandon Rhodes | 08 February 2015, 17:12:25 UTC | Title-case target names when printing a segment | 08 February 2015, 17:12:25 UTC |
19248ff | Brandon Rhodes | 08 February 2015, 16:58:35 UTC | Fix D''ARREST so that it becomes D'ARREST | 08 February 2015, 16:58:35 UTC |
95199b0 | Brandon Rhodes | 08 February 2015, 10:32:37 UTC | Start documenting new SPK capabilities | 08 February 2015, 10:32:37 UTC |
f7cf0bf | Brandon Rhodes | 08 February 2015, 09:29:00 UTC | Remove legacy ephemerides instructions from README | 08 February 2015, 09:29:00 UTC |
e9896e4 | Brandon Rhodes | 08 February 2015, 09:27:09 UTC | Quick benchmark script to test new performance | 08 February 2015, 09:27:09 UTC |
360ed01 | Brandon Rhodes | 08 February 2015, 09:18:49 UTC | Switch to SPK.open() if the user has a path Because, what if they have an open file instead? The previous design made it impossible to ever just keep their file open and have the SPK class use it. | 08 February 2015, 09:18:49 UTC |
f928bbe | Brandon Rhodes | 08 February 2015, 08:52:30 UTC | Elide now-useless "segment" parameter to compute() | 08 February 2015, 08:52:30 UTC |
1c49409 | Brandon Rhodes | 08 February 2015, 08:40:13 UTC | Add claims of Python 2.6 and 3.4 compatibility | 08 February 2015, 08:40:13 UTC |
7a4e856 | Brandon Rhodes | 08 February 2015, 08:37:39 UTC | Remove extraneous "print" from test | 08 February 2015, 08:37:39 UTC |
f339fbd | Brandon Rhodes | 08 February 2015, 08:37:04 UTC | Be more careful to make array dimensions ints | 08 February 2015, 08:37:04 UTC |
2491588 | Brandon Rhodes | 08 February 2015, 08:31:53 UTC | Switch JPL tests to our SPK code with DE430 | 08 February 2015, 08:31:53 UTC |
9447048 | Brandon Rhodes | 08 February 2015, 08:15:45 UTC | Have segments, when printed, include object names | 08 February 2015, 08:15:45 UTC |
bc8192f | Brandon Rhodes | 08 February 2015, 07:19:00 UTC | Quick awk script to parse target names | 08 February 2015, 07:19:00 UTC |
6e64e93 | Brandon Rhodes | 08 February 2015, 06:52:23 UTC | Make SPK pretty-print, and able to return scalars Two conveniences that I myself need while working with "jpltest.py". | 08 February 2015, 06:52:23 UTC |
79b5343 | Brandon Rhodes | 08 February 2015, 06:31:39 UTC | Remember to tell Travis where package lives now | 08 February 2015, 06:31:39 UTC |
23f9f7f | Brandon Rhodes | 08 February 2015, 06:29:32 UTC | Move crucial DAF format checks earlier during init Checking the file type should be first, before trying to determine the endian-ness. | 08 February 2015, 06:29:32 UTC |
fff52d0 | Brandon Rhodes | 08 February 2015, 05:40:43 UTC | Promote "jplephem" and its "setup.py" to top level Now that the main point of this package is the "jplephem" module for reading SPK kernels, we should go ahead and promote it to being a top level directory. | 08 February 2015, 05:40:43 UTC |
5d5d52b | Brandon Rhodes | 08 February 2015, 05:37:51 UTC | Remove old "tox.ini" file | 08 February 2015, 05:37:51 UTC |
a56040b | Brandon Rhodes | 08 February 2015, 05:32:11 UTC | Move old Python-format ephemerides to "legacy" Because bare directories with *.py files inside are now considered packages to Python 3, these directories were being grabbed by the "import" statement during certain tests, instead of letting Python find the actual installed ephemerides packages. Bumping them down a level solves the problem, and also unclutters the top level. | 08 February 2015, 05:32:11 UTC |
d8180b7 | Brandon Rhodes | 08 February 2015, 05:31:42 UTC | Git-ignore files of test data from JPL | 08 February 2015, 05:31:42 UTC |
b45eacc | Brandon Rhodes | 08 February 2015, 05:15:25 UTC | Write, and test, a pretty str() for Segment | 08 February 2015, 05:15:25 UTC |
1cc4d8b | Brandon Rhodes | 08 February 2015, 05:03:20 UTC | Repent of crazy namedtuple; make Segment a class | 08 February 2015, 05:03:20 UTC |
75ff738 | Brandon Rhodes | 08 February 2015, 04:05:00 UTC | Remove dict comprehension to succeed under 2.6 | 08 February 2015, 04:05:00 UTC |
c9a51f1 | Brandon Rhodes | 08 February 2015, 02:30:24 UTC | Fix Travis tests by downloading necessary file | 08 February 2015, 02:30:24 UTC |
e621443 | Brandon Rhodes | 08 February 2015, 01:48:05 UTC | Add SPK to tests, and clean up SPK some more | 08 February 2015, 01:48:05 UTC |
7d5d843 | Brandon Rhodes | 08 February 2015, 00:31:42 UTC | Parameterize tests to be independent of Ephemeris | 08 February 2015, 00:31:42 UTC |
ff46a58 | Brandon Rhodes | 08 February 2015, 00:26:23 UTC | Tighten most test precision by 100x | 08 February 2015, 00:26:23 UTC |
e800059 | Brandon Rhodes | 08 February 2015, 00:21:29 UTC | Try to make tests importable under Python 2.6 | 08 February 2015, 00:21:29 UTC |
27a7734 | Brandon Rhodes | 08 February 2015, 00:16:48 UTC | Add Travis CI integration | 08 February 2015, 00:16:48 UTC |
541302e | Brandon Rhodes | 08 February 2015, 00:16:26 UTC | Skip simple tests if DE421 is not installed | 08 February 2015, 00:16:26 UTC |
d5e0de7 | Brandon Rhodes | 08 February 2015, 00:08:46 UTC | Ignore SPK files downloaded for tests | 08 February 2015, 00:08:46 UTC |
debe227 | Brandon Rhodes | 07 February 2015, 17:56:38 UTC | Add a docstring to the comments() method | 07 February 2015, 17:56:38 UTC |
19a4c3e | Brandon Rhodes | 07 February 2015, 17:49:00 UTC | Add DAF comments() method for reading comment area | 07 February 2015, 17:49:00 UTC |
5c07040 | Brandon Rhodes | 04 February 2015, 13:01:24 UTC | Give SPK object its own Chebyshev logic | 04 February 2015, 13:01:24 UTC |
a85b7de | Brandon Rhodes | 04 February 2015, 08:41:27 UTC | Streamline and simplify DAF and SPK parsing | 04 February 2015, 08:41:27 UTC |
c1c3edb | Brandon Rhodes | 04 February 2015, 08:06:57 UTC | Split out DAF and SPK classes into separate files | 04 February 2015, 08:06:57 UTC |
401a3b2 | Brandon Rhodes | 04 February 2015, 08:01:57 UTC | Remove initial experiment in striding NumPy array | 04 February 2015, 08:01:57 UTC |
23ba549 | Brandon Rhodes | 04 February 2015, 08:00:05 UTC | Move "daf.py" into place inside "jplephem" | 04 February 2015, 08:00:05 UTC |
5f1348f | Brandon Rhodes | 04 February 2015, 07:57:38 UTC | Make predictions exact by fixing time endpoints | 04 February 2015, 07:57:38 UTC |
fcbc9e2 | Brandon Rhodes | 04 February 2015, 07:26:24 UTC | Coax SPK routine into a (mostly wrong) computation | 04 February 2015, 07:26:24 UTC |
a34821b | Brandon Rhodes | 04 February 2015, 06:23:27 UTC | Start designing a new Ephemeris subclass for SPK | 04 February 2015, 06:23:27 UTC |
6660a1f | Brandon Rhodes | 04 February 2015, 05:59:43 UTC | Repair code so it runs under Python 2 | 04 February 2015, 05:59:43 UTC |
2cc458e | Brandon Rhodes | 04 February 2015, 05:42:05 UTC | Merge successful experimental repo 'python-spk' | 04 February 2015, 05:42:05 UTC |
ac02eb9 | Brandon Rhodes | 03 February 2015, 09:43:32 UTC | A triumph: wrap coefficients in-place in ndarray! | 03 February 2015, 09:43:32 UTC |
fbc3f65 | Brandon Rhodes | 03 February 2015, 09:16:22 UTC | Pull out radius and midpoint values as arrays | 03 February 2015, 09:16:22 UTC |
40b2a3e | Brandon Rhodes | 03 February 2015, 08:57:40 UTC | Make code compatible with Python 3 | 03 February 2015, 08:57:40 UTC |
ff68cb6 | Brandon Rhodes | 03 February 2015, 08:51:02 UTC | Interpret dimensions of SPK coefficient set | 03 February 2015, 08:51:02 UTC |
27b3a2d | Brandon Rhodes | 02 February 2015, 09:34:11 UTC | Provide a way to iterate across all summary data | 02 February 2015, 09:34:11 UTC |
6d771f7 | Brandon Rhodes | 02 February 2015, 09:03:32 UTC | Correctly parse the first summary in the file | 02 February 2015, 09:03:32 UTC |
8a9bc1c | Brandon Rhodes | 02 February 2015, 08:22:03 UTC | Rename to "daflib.py" to avoid name conflicts | 02 February 2015, 08:22:03 UTC |
4569801 | Brandon Rhodes | 02 February 2015, 08:21:48 UTC | Switch from slow file I/O to an mmap | 02 February 2015, 08:21:48 UTC |
d3950cd | Brandon Rhodes | 02 February 2015, 07:02:30 UTC | First steps toward parsing DAF file record header | 02 February 2015, 07:02:30 UTC |
8b26d8a | Brandon Rhodes | 11 June 2014, 15:52:58 UTC | Make "refetch.sh" survive "./tmp" already existing | 11 June 2014, 15:52:58 UTC |
33c82c8 | Brandon Rhodes | 11 June 2014, 15:51:57 UTC | Remove specific Python version numbers These packages are intended to remain evergreen while new Python versions come out, so I am keeping the Python 2 and Python 3 compatibility assertions but removing any more specific version assertions from the trove classifiers. | 11 June 2014, 15:51:57 UTC |
92b06c0 | Brandon Rhodes | 06 January 2014, 18:41:49 UTC | Add links from README to the 5 ephemeris packages | 06 January 2014, 18:41:49 UTC |
49c189a | Brandon Rhodes | 06 January 2014, 17:50:24 UTC | Add project repository URL to ephemeris setup.py's | 06 January 2014, 17:50:24 UTC |
dfab730 | Brandon Rhodes | 06 January 2014, 17:40:03 UTC | Update version numbers to end in ".1"; remove URLs This is necessary for recent versions of pip to consider these packages production instead of beta, and thus be willing to install them. Also, this patch removes all old URL fields since we will now need to upload these packages directly to PyPI, since PyPI by default no longer pulls packages from 3rd party locations. Also, this removes all detritus that was left over in the setup.py files from the era back when I just had a single combined setup.py that was shared by all of the packages. | 06 January 2014, 17:40:03 UTC |
659cdbc | Brandon Rhodes | 04 January 2014, 19:27:53 UTC | Further improve the wording about JD precision | 04 January 2014, 19:27:53 UTC |
0fa95bf | Brandon Rhodes | 04 January 2014, 19:20:54 UTC | Add link to USNO 64-bit Julian Date paper | 04 January 2014, 19:20:54 UTC |
b7dad2a | Brandon Rhodes | 26 November 2013, 22:55:53 UTC | Release 1.2 | 26 November 2013, 22:55:53 UTC |