lcogt / banzai-floyds Goto Github PK

Description:

We need to reliably capture and extract the FLOYDS orders.

Parameters for Success:

• Capture >90% of pixels in both lit orders
• Less than 1% of order pixels from background outside of order
• Orders successfully differentiated from each other.

Tasks:

• Create Synthetic orders with known positions
• Plot order overlay for sanity check
• Manually create mask of orders for comparison with real data (Lamp, Arc and science)
• Compare to current pipeline order extraction

Description:

We need to have a fully documented pipeline so that both users and developers know exactly what the pipeline is doing and how to run it themselves.

Parameters for Success:

3rd party must be able to install, run, and understand pipeline based on provided documentation.

Tasks:

• Installation Documentation
• Run Documentation
• Order Extraction Documentation
• Trace Extraction Documentation
• Wavelength Calibration Documentation
• Fringe Removal Documentation
• Flux Calibration Documentation
• Telluric Removal Documentation
• Update lco.global, archive old pipeline docs

Description:

We need to understand how well we extract telluric lines from from the trace.

Parameters for Success:

• 90% of telluric lines should be removed

Tasks:

• Compare to featureless spectrum (White dwarf? D-type asteroid?)
• Plot results for Sanity Check

Description:

We need to be able successfully extract the trace from an order and remove the background from the data.

Parameters for Success:

• Successfully detect and extract a trace with a S/N of >10 above background.

Tasks:

• Create synthetic trace
• Implement visualizations for sanity check
• Manually mask out traces on real data for comparison
• Compare to current Pipeline

Description:

We need to understand how well we can calibrate the spectral flux compared to standard stars.

Parameters for Success:

• Flux should be reliably calibrated across entire spectrum within 2 sigma

Tasks:

• Calibrate one standard using another
• Plot comparison results for a sanity check.

Description:

We need to understand how BANZAI FLOYDS data products compare to the data products from the current FLOYDS pipeline. This will not only give us a metric to determine the success and validity of the BANZAI pipeline, but also allow us to better understand differences and improvements between the two.

Parameters for Success:

We should provide AT LEAST the same quality data as provided by the current pipeline.
We should carefully consider the output data products to ease the transitional burden off of users where possible.
We should test a variety of different kinds of targets and configurations against the old pipeline.

Tasks:

• Compare synthetic data run through both pipelines
• Compare Standards run through both the current pipeline and Banzai-FLOYDS
• Compare Science Data run through both the current pipeline and BANZAI FLOYDS

Description:

All of the above tests should work for a variety of target brightnesses and sky conditions

Parameters for Success:

• All above parameters are met for a variety of target conditions.

Tasks:

• Test for bright targets (<8 Mag)
• Test for Faint targets (>15 Mag)
• Test for Bright sky (Full Moon)
• Test for moving target
• Test for unusual slit positions

Description:

We need to understand how well we can remove fringing from the red side of the spectrum.

Parameters for Success:

• Remove 90% of fringing visible in the red.

Tasks:

• Quantify the amount of fringing using a normalized smoothing function (median value, i.e.) in the proper region as compared to the variance in that region.
• Quantify 0 fringing by applying the above method on the blue arm of the spectrum
• Compare across many different targets, temperatures, and slit widths
• Use PKs Visualizations for sanity check

Description:

All of the above tasks should function at all slit widths.

Parameters for Success:

• Above parameters are met for all slit widths.

Tasks:

• Test on 1.2" slit
• Test on 1.6" slit
• Test on 2.0" slit
• Test on 6.0" slit

Description:

We need reliable end to end tests that quickly and efficiently test that every stage of the pipeline is in good working order.

Parameters for Success:

1. Tests run successfully
2. All Stages are tested
3. Expected products are created
4. Works for a variety of actual science frames

Tasks:

• [ ]

Description:

We need to understand how precisely and reliably we can calibrate the wavelength of a spectrum

Parameters for Success:

• Should be accurate to within 1 Angstrom.

Tasks:

• Create synthetic spectra
• Build Visualization options for a sanity check
• Compare to target with known bright, narrow lines.
• Compare to Current pipeline wavelength calibration