Closed kapadia closed 10 years ago
Are there any plans of implementation of this? Is there a clear benefit than the current form? I'd like to see the what's next on your list for this project. :)
@Merg1255 This is an old issue that was never closed. It's implemented here:
https://github.com/astrojs/fitsjs/blob/master/src/tabular.coffee#L67
but I'd like to optimize for huge tables (> 100,000 rows). At the moment, it's inefficient because an array (as opposed to a fixed length typed array), is storing the column values.
https://github.com/astrojs/fitsjs/blob/master/src/tabular.coffee#L92
I have some ideas to improve it, but in it's current state it works well for tables with < 100,000 rows. The API will stay the same regardless of any optimizations.
Good to know. I'll be watching for updates. :)
@Merg1255 getColumn is now an order of magnitude faster. Was able to get a column containing 3.3 million rows in ~22 seconds.
@kapadia that's great!! :)
Get column should be implemented such that bytes are skipped appropriately to avoid unneeded calls to DataView getters methods.