Replacing the IDL AstroLIB functions?JythonBuilding a replacement for numpyNot compatible with numpyUse JNI to pull C components of numpy into Jython?JNI is platform dependentJNI code would be hard to distributeJython projects "need" to be
We have already switched from numarray to numpyBooklet on doing the conversionData analysis tutorial has been updated to use numpyNeed a numpy reference that is more than a tutorial but less than Travis' bookNeed to make our internal documentation more consistentNeed a reference document for our productsNeed a matplotlib cookbook geared for astronomerssimple "how-tos" for generating plotscommon examples and some offbeat things that are hard to figure outMaking publication quality plots
Spitzer SoftwarePossible replacement for IRAF?Very genericNot compatible with existing IRAF tasksIs anyone interested in helping out?Written in PythonThink of it as an IDE for data reductionBased upon IRAS extraction routinesCurrent system has a plugin architectureLicensed under the GPLEverything is in PythonUses qt and pyqt for interface codeAll file management is done internallyIdeally you would only need a basic knowledge of the frameworkNo work has been done on the photometric side of thingsCurrently operates for spectral extractionCommentsQuick look niche application?Not likely used for publication quality data reductions?Thing to do would be to put it out there and see who uses itCurrently not very public
How to communicate to astronomy community?Astropy?not high traffic -- STScI's faultmore of a developer listScipy, numpy listsci.astro the groupused by observers
Performance IssuesTablesHeader ManipulationsAdded featuresstreaming data modecode has been reorganizedallows for STScI conventionsRead FITS files given a URLCan read zipped filesSupport for compressed extensionsInterested in parallel computingCould pyfits read and write to streams?we would need to look into itSome scaling cases we don't return the correct "type" for the objectint32 to float32 will lose person precisionwe will likely just return a higher precisionVOTable support
Driven by GeminiCurrently fixing bugs and adding featuresRemove the dependence on IRAFAdd matplotlib as a graphics interface
We pull things out of scipy so that we don't have such a large dependencyFix the installation issueseggs?The numarray packages now live under scipyndimageimageconvolveWe can add some of the functionality that lives in IDL to scipyFor scipy 1.0 releaseneed docsdistribution issueinstallation of individual packages
we are actively working on fixing holes we have foundmath textmemory leaks in GUIsdistribution issuesneeds to be easierbundle dependent libraries?zilb, freetypemake setup smarterIs there a way to fix an image size easily?Yestry fig imageraw pixel dumpno interpolationMake heavier use of traits?Changes to backend modules?Merge with Chaco?If serious refactoring of internals then the object interface will likely break. The pylab interface should remain the same.
STScI wants to do more workdone coordinate workSomeone posted a wrap for WCS supportWorking on a python approach to synphoton astrolib svn but not advertisedActivity from STScI should pick up againSome people interested in doing literal translations of IDL aston utilities
Replacing IRAF tasks with Python versionsPhotometric packagesdaophotapphotTake current ongoing projects and make them Python friendlywould not be rewritesWhat is going to replace IRAF?Parallel processingHow can we use multicore processors that is not too onerousit seems easy to run Python tasks as separate processesmemory mapping as a way of sharing memory?Threads can't alter reference counts arbitrarilyGlobal Interpretor Lock allows only one thread to run at a timeIs there going to be a replacement to skycat?All replacements seem to have gone unsupportedThere is a need for something that fills skycat's role