need report of tuning OK/not OK on console
current plan indicates success, even if axis did not tune

Must work with both EpicsScaler and ScalerCH devices

# will become part of APS_BlueSky_tools.devices
def use_EPICS_scaler_channels(scaler):
    """
    configure scaler for only the channels with names assigned in EPICS 
    """
    read_attrs = []
    for ch in scaler.channels.component_names:
        _nam = epics.caget("{}.NM{}".format(scaler.prefix, int(ch[4:])))
        if len(_nam.strip()) > 0:
            read_attrs.append(ch)
    scaler.channels.read_attrs = read_attrs

running this code:

In [3]:     APS_devices.AD_setup_FrameType(
   ...:         EPICS_PV_prefix["PG3 PointGrey Grasshopper3"], 
   ...:         scheme="DataExchange"
   ...:     )

produces this exception:

~/Apps/BlueSky/lib/python3.6/site-packages/APS_BlueSky_tools/devices.py in AD_setup_FrameType(prefix, scheme)
    840     template = "{}cam1:FrameType{}.{}"
    841     for field, value in db.items():
--> 842         epics.caput(template.format(prefix, "", field), value)
    843         epics.caput(template.format(prefix, "_RBV", field), value)
    844 

NameError: name 'epics' is not defined

based on verbal discussions with NSLS-II DAMA team, best to report this as custom event stream

reading information from Excel spreadsheets is a major plus

bring in the routines developed for building the SAS2018 Conference Book

print() is used sometimes for diagnostics and debugging

Use python's logging package instead

the open() and close() methods should rely on set("open") and set("close") to work properly in a plan and on command line

see APS-2BM-MIC/ipython-user2bmb#22

zmq_pair

Create a plan to scan a pair of axes in a lissajous trajectory. This has advantages over snake or typewriter scans each time under certain limitations. When reducing exposure of the sample to the beam is important, the snake or typewriter scan increase the sample exposure during acceleration and deceleration of the inner axis. In such scans, only one axis moved at a time. The inner axis traverses its range, then the outer axis is incremented.

In a lissajous scan, both axes move simultaneously along trajectories that are functions of time, each with a different frequency. There is an adjustable phase difference applied that remains constant through the scan.

With judicious choices of frequency for each axis, sampling density rivaling the snake or typewriter scan should be possible.

Ptychography is one particular application that comes to mind which could benefit from a lissajous scan.

references

• https://en.wikipedia.org/wiki/Lissajous_curve
• http://mathforum.org/mathimages/index.php/Lissajous_Curve

This line in PeakAxisTuner.tune():

    self.peak_stats = PeakStats(x=self.axis.name, y=self.det.name)

should allow for the signal to be named other than self.det.name.

For example, a signal from the swait record support might pick:

    self.peak_stats = PeakStats(x=self.axis.name, y=self.det.val.name)

The APS delivers X-ray beam in operations mode. At other times, the storage ring may have significant electron current but the beam is not available for X-ray users. The month-long maintenance periods, occurring three times yearly, are examples. It is still useful to use these periods for code development and maintenance. When beam lines add suspenders for low storage ring current or front end shutter closed, it would be useful to only add those when in operations. During a major shutdown, a beam line might implement a shutter software simulator in place of the physical controls to facilitate code maintenance.

Provide a boolean property function to declare if the APS is in an operations period as a method of the devices.ApsMachineParametersDevice() class.

zmq_pair

zmq_pair

see APS-USAXS/usaxs-bluesky-ended-2023#12

time to create general documentation (out of source code) with examples and figures and stuff

here's a checklist to clear ... created from items in the RTD sidebar

• Demo
• Callbacks
• Demonstrate nscan and TuneAxis
• Devices
• Examples
• File Writers
• Plans
• bluesky_snapshot
• sscan record
• Suspenders
• swait record
• Utilities
• command-line applications

this is published for conda on my channel now

The XIA PF4 filter is supported in EPICS synApps. Provide ophyd device support. The most common use at the APS is in the dual-PF4 box combination.

ophyd.EpicsMotor device lacks:

• .DESC description
• .DVAL & .DRBV : dial coordinates
• .RVAL & .RRBV : raw coordinates
• .CNEN : enable servo mode

ophyd.EpicsMotor does support .FOFF as offset_frozen and .OFF as user_offset

As discussed in ophyd issue 598, test if the area detector instance is ready for data acquisition with Bluesky. If it is not ready, a TimeoutError exception will be raised when the HDF5 plugin is triggered.

This may not be specific to the HDF5 plugin. Other plugins might need similar tests.

Users will write and call procedural functions (non-data collection) for use during their plans and these functions are likely to have blocking calls. Provide support so these functions can be run in a separate thread. Requirements are that each function takes no arguments and must not return until its activities are complete.

zmq_pair

following APS-2BM-MIC/ipython-user2bmb#42

all shutters should have a boolean property to report if shutter is open or closed

see APS-USAXS/usaxs-bluesky-ended-2023#27

The users want to be able to control the area detector file name and are not happy being treated like idiots. Give them the ability to shoot themselves in the foot by allowing them to control the file name in bluesky plans.

storage ring current, ACIS permit, operator messages, ... everybody will want some of these - define them in a standard device

• python package
• PyPI project
• conda package
• sphinx documentation
• unit testing
• travis-ci

Some beam lines are using a feature of the EPICS Area Detector HDF5 File Writer plugin to specify the type of frame being acquired by the detector, whether image, dark, white (flat), or possibly something else.

Provide a way to configure this for general use.

A Jupyter notebook has been created to describe the scheme: https://github.com/BCDA-APS/use_bluesky/blob/master/notebooks/images_darks_flats.ipynb

Make a client receiver for the ImageZMQCallback() that identifies image data and rehsapes into numpy image arrays.

Can prove this by creating PNG files.

these newer document might be use for console or other reporting (such as comments in SPEC file)

In [9]: RE(bp.count([alta_det]))
Transient Scan ID: 204     Time: 2018/07/27 09:43:07
Persistent Unique Scan ID: '9355a621-6067-4ab6-83d9-32c200c6afb7'
New stream: 'primary'                                                                                                                                                       
+-----------+------------+
|   seq_num |       time |
+-----------+------------+
custom_callback encountered: resource : {'spec': 'AD_HDF5', 'root': '/', 'resource_path': 'share1/USAXS_data/test/alta/2018/07/23/pilgrim_000000.h5', 'resource_kwargs': {'frame_per_point': 0}, 'path_semantics': 'posix', 'uid': 'eb6d3c38-2008-49e9-a90c-6c6cca8f08d7', 'run_start': '9355a621-6067-4ab6-83d9-32c200c6afb7'}
custom_callback encountered: datum : {'resource': 'eb6d3c38-2008-49e9-a90c-6c6cca8f08d7', 'datum_id': 'eb6d3c38-2008-49e9-a90c-6c6cca8f08d7/0', 'datum_kwargs': {'point_number': 0}}
|         1 | 09:43:09.4 |
+-----------+------------+
generator count ['9355a621'] (scan num: 204)



exit status: success
# descriptor(s): 1
# event(s): 1
Out[9]: ('9355a621-6067-4ab6-83d9-32c200c6afb7',)

USAXS likes to set motor soft limits for various modes

Provide a mixin to access HLM & LLM fields, with some compatibility methods for get_lim and set_lim

This code, developed for the MONA project, is no longer needed or used. No point in keeping it.

zmq_pair

relates to #23 and to bluesky/bluesky#997

Count time is implemented differently, based on the detector type (and conceivable other hardware-specific features). Here's the ones I know about now:

• SynSignal : bps.mv(det.exposure_time, time))
• EpicsScaler : bps.mv(det.preset_time, time)
• ScalerCH : bps.mv(det.preset_time, time)
• areadetector.DetectorBase : bps.mv(det.cam.acquire_time, time)

Supporting negative time to count against a monitor detector is only supported by EpicsScaler and ScalerCH in the underlying EPICS record support. That may have to drop from this feature request.

see APS-USAXS/usaxs-bluesky-ended-2023@45ed8e9

this already seems to work. Might just dump APS_BlueSky_tools.plans import TuneAxis code

zmq_pair, consequence of #14, #15, #16, #17

How to get info (and/or data) from each new frame into BlueSky stream?

Current ophyd support was built around earlier AD version. The API has changed and that support fails.

see: http://readthedocs.org/projects/aps-bluesky-tools/builds/6359040/
These errors block the documentation of plans.py

ImportError: No module named zmq
ImportError: No module named ophyd.device

zmq_pair

Examine the existing data collection procedures used by APS tomography for algorithm and metadata. Incorporate where possible in #8.

zmq_pair

The data does not get into PeakStats and no peak is found. Channel names are different than EpicsScaler

zmq_pair

As stated for the APS USAXS instrument, any axis should be able to have bespoke tuning, via a call to its self.axis.tune() method. The tune() method will be provided by a Mixin class which also provides an additional attribute tuner (a subclass of AxisTunerBase) that defaults to a standard tuner instance but may be changed to one customized for the axis.

A skeleton of this idea (for APS_BlueSky_tools.devices) is:

class TuningException(ValueError): 
    """Exception during execution of `AxisTunerBase` subclass"""
    pass

class AxisTunerBase(object):
    ok = False
    center = None
    def tune(md=md, **kwargs): ...
    def config(**kwargs): ...
    def peak_detected(): ...

class PeakAxisTuner(AxisTunerBase):
    """
    scan y v. x, then set x = x@max(y), iff max(y) >= 4*min(y), raise `TuningException` as needed
    """
    # implement algorithm from SPEC here (tune to signal peak)
    # see example: https://github.com/prjemian/ipython_ookhd/blob/master/profile_bluesky/startup/11-bs_sim_motors.py#L29
    pass

class Mixin(object):
    # Mixin MUST not provide __init__() method, instead use self.tuner.config()
    tuner = PeakAxisTuner()
    def tune(md=md, **kwargs):
        # TODO: prep
        if self.tuner is not None:
            yield from self.tuner.tune(md=md, **kwargs)
        # TODO: restore

zmq_pair

from the SpecWriterCallback() filewriter:

/APSshare/anaconda3/BlueSky/lib/python3.6/site-packages/APS_BlueSky_tools/filewriters.py in receiver(self, key, document)
    211             msg = "custom_callback encountered: {} : {}".format(key, document)
    212             # raise ValueError(msg)
--> 213             logger.warn(msg)
    214         return
    215 

UnboundLocalError: local variable 'logger' referenced before assignment

In [4]: !geany /APSshare/anaconda3/BlueSky/lib/python3.6/site-packages/APS_BlueSky_tools/filewriters.py

Before the rotStage is moved, it's current position is set to rotStage.position%360.0. This needs to be done with the .OFF field frozen (.FOFF=1).

Add newfile() command to SPEC writer class and make it write more than one scan per file. Also need to support re-opening an existing SPEC file.

• pass file name (not path) when creating the instance, default to date/time file name in pwd if file name not provided
• What about reset the scan_id with a newfile()?