Using PlateMaker in the EM challenge

Origin:$HeadURL: https://desi.lbl.gov/svn/code/online/PlateMaker/trunk/doc/calibproc.rst $ Revision:$Rev: 4907 $ Date:$Date: 2016-06-08 16:13:50 -0500 (Wed, 08 Jun 2016) $ Author:$Author: neilsen $

Preparing the test environment

Cleaning the working directory

The examples here assume a clean working directory, which needs to be set by the data_dir parameter in the PLATEMAKER role definition in the example platemaker.ini file, found in $PLATEMAKER_DIR/etc.

Make sure the parent directory exists:

>>> from os import mkdir
>>> try:
...     mkdir('/tmp/test_platemaker_data')
... except FileExistsError:
...     pass

but the test subdirectory does not. (PlateMaker generates it automatically.):

>>> from shutil import rmtree
>>> try:
...     rmtree('/tmp/test_platemaker_data/52256')
... except FileNotFoundError:
...     pass

Loading test input data

These examples require data to be provided to PlateMaker. These can be loaded from pickles in the PlateMaker product:

>>> from PlateMaker.PlateMakerTestData import PlateMakerTestData
>>> test_data = PlateMakerTestData('em2', 52256)
>>> seq_id = test_data.seq_id
>>> positioners = test_data.positioners
>>> merged_fvc_spots = test_data.merged_fvc_spots

Preparing to interact using PML

The architect with a PlateMaker role needs to be up and running, and the environment variables need to be set up so that the python process can find it, for example with dos_env.csh.

Then, start a python interactive process and establish a connection to the PlateMaker role:

>>> from DOSlib.PML import dos_connection
>>> pm = dos_connection('PLATEMAKER')

Running calibprep

The first step in running calibprep is to set the instrument. In this example, I begin by setting parent directory for the instrument definition files:

>>> pm.execute('configure')
'SUCCESS'
>>> pm.execute('set', instrument_dir=test_data.instrument_dir)
'SUCCESS'

This value can also be set in the PlateMaker section of the architect configuration file; if it is set there, setting it as above is unnecessary.

Once PlateMaker is told where to look for instrument definitions, tell it what instrument to use:

>>> pm.execute('set', instrument='em2')
'SUCCESS'

Now you can set the sequence id, seq_id:

>>> pm.execute('set', seq_id=seq_id)
'SUCCESS'

calibprep can be run thus:

>>> pm.execute('set', positioners=positioners)
'SUCCESS'
>>> pm.execute('calibprep')
'SUCCESS'

Note that the set and the calibprep executions can be combined into a single call:

>>> pm.execute('calibprep', positioners=positioners)
'SUCCESS'

After calibprep is run, the data can be retrieved thus:

>>> fiber_pos = pm.execute('get', 'fiber_pos')

Running calibproc

calibproc can be run thus:

>>> pm.execute('calibproc', positioners=positioners,
...                         merged_fvc_spots=merged_fvc_spots)
...
'SUCCESS'

Again, when run after prior commands which have provided the necessary data, the previously provided keywords may be omitted:

>>> pm.execute('calibproc', positioners=positioners, merged_fvc_spots=merged_fvc_spots)
'SUCCESS'

After calibproc is run, the data can be retrieved thus:

>>> positioner_calib = pm.execute('get', 'positioner_calib')