generalize the tomo_scan() and user_tomo_scan() plans

[prjemian]

These working plans have protocols that are general but also have protocols that are very specific to the instrumentation in 2-BM-B. Generalize for use in other stations and beam lines (32-ID-C comes to mind). Ultimately, tomo_scan() might become part of a standard library (perhaps APS_BlueSky_tools) for general use.

[prjemian]

related to #48

[prjemian]

Summary of 2018-08 Taxi/Fly Tomo Scan

• my_series(start, stop, step, acquire_time=0.02):

  • runs user_tomo_scan at series of tomo_stage.y positions

• user_tomo_scan(*, acquire_time=0.02, iterations=1, delay_time_s=1.0, samOutDist=-3, md=None):

  • setup:
    • tomo rotary stage
    • tomo scan
    • tomo detector
    • metadata
  • repeat for specified iterations:
    • prepare:
      • detector (and image acquisition)
      • rotary stage
      • shutter
      • tomo_scan:

• tomo_scan(*, start=0, stop=180, numProjPerSweep=1500, slewSpeed=5, accl=1, samInPos=0, samOutDist=-3, acquire_time=0.02, md=None):

  • setup:
    • tomo rotary stage
    • tomo scan
    • tomo detector
    • metadata
    • additional signals to be monitored
    • interim progress reporting
    • cleanup
  • scan:
    • prepare:
      • monitors
      • rotary stage
    • measure darks and flats
    • open shutter
    • taxi
    • fly
  • after the scan is done:
    • read detector and save data to primary stream
    • update HDF5 file with rotary table position array

• cleanup:

  • removes monitors:
    • rotary stage
    • detector
  • resets taxi/fly
  • restores default settings
    • detector
    • rotary stage
    • signals from supervisory processes (such as MONA)
  • closes shutter
  • returns rotary stage to default position

[prjemian]

rough taxi/fly algorithm for tomo

repeat
    prepare
    measure darks
    measure flats
    prepare for tomo
    taxi
    fly
    collect
    clean up

[prjemian]

rough step-scan algorithm for tomo

repeat
    prepare
    measure darks
    measure flats
    prepare for tomo
    repeat
        rotate
        trigger
        collect
    clean up

[prjemian]

Consensus of the NSLS-II DAMA team is for bluesky plans to be functions rather than class methods. For tomography scans, it seems more natural in Python to create a strong base class with subclasses for step and taxi/fly, then subclasses for each instrument.

[prjemian]

@decarlof : Are you comfortable with the intention to create a generalized TomographyScanBase(object) base class with subclasses for step and taxi/fly, then subclasses for each instrument. The class will have placeholders for the rotation axis, detector, shutter, displacement axis (for white frames), ... and a bluesky plan comparable to tomo_scan() above. The user_tomo_scan() (common settings picked out by the instrument team) would be defined in the instrument subclass.

import APS_BlueSky_tools.plans as APS_plans
class TomographyScanBase(object)
    def tomo_scan(self, ...)
class TomographyStepScan(TomographyScanBase)
    def tomo_scan(self, ...)
class TomographyTaxiFlyScan(TomographyScanBase)
    def tomo_scan(self, ...)

# at 2-BM-B
class TomographyTaxiFlyScan2BMB(APS_plans.TomographyTaxiFlyScan)
    def tomo_scan(self, ...)

[prjemian]

First draft in another repo: https://github.com/BCDA-APS/APS_BlueSky_tools/blob/master/examples/tomo_issue49.ipynb

[prjemian]

Incorporate #46 (add delay between each dark and flat).

[prjemian]

Incorporate #48 (detector-specific info should be part of detector object)