Content
Intro
This page describes dynamic generation of gain correction calibration constants for multi-panel jungfrau detector. This is relevant to calibration types pixel_gain and pixel_offset
pixel_gain: shape:(3, 1, 512, 1024) size:1572864 dtype:float32pixel_offset
which are combined from 3 gain range per panel arrays of shape (512, 1024).
Naming conventions
Since psana.Jungfrau.ConfigV3 software provides per-segment access to the class ModuleConfigV1 object with three methods
- moduleVersion()
- firmwareVersion()
- serialNumber()
which return 64-bit integer numbers like 1511699, 23403815178277, 3997947, respectively. However, some of these numbers in hexadecimal format looks as dates and times. For better readability the panel Id name can be formatted as '%s-%s-%s', for hexadecimal values with dropped '0x' as hex(v).lsplit('0x'), e.g. "171113-154920171025-3d00fb".
Then, total Jungfrau Id is a variable-length underscore-separated concatenation of sequential panel Id names, i.g. for two panels of exp=xcsls3716:run=631, XcsEndstation.0:Jungfrau.0 complete detector Id is 170505-149520170815-3d00b0_170505-149520170815-3d00f7.
API
Method returning Jungfrau and its segment Ids is implemented in Detector/src/UtilsJungfrau.py and can be used as in example:
import psana
from Detector.UtilsJungfrau import id_jungfrau
ds = psana.DataSource('xpptut15:run=430')
env = ds.env()
src = 'MfxEndstation.0:Jungfrau.0' # or part of the DetInfo name 'Jung' or 'Jungfrau.1' or alias 'jungfrau1M'
print 'id_jungfrau(env, src, 0):', id_jungfrau(env, src, iseg=0) # for segment 0
print 'id_jungfrau(env, src, 1):', id_jungfrau(env, src, iseg=1) # for segment 1
print 'id_jungfrau(env, src) :', id_jungfrau(env, src) # for all segments
which prints
id_jungfrau(env, src, 0): 171113-154920171025-3d00b0 id_jungfrau(env, src, 1): 171113-154920171025-3d00f7 id_jungfrau(env, src) : 171113-154920171025-3d00b0_171113-154920171025-3d00f7
CLI
A few commands which shed light on data with jungfrau detector are listed below.
Command line to get jungfrau psana.EventKey objects:
event_keys -d exp=xpptut15:run=430 -m2 -p Jung
Command line to find experiment runs with jungfrau:
find_detector_runs MFX jungfrau
Command line to get jungfrau Id (in release>ana-1.3.51):
jungfrau_id <=== prints help jungfrau_id exp=xpptut15:run=430 [Jungfrau.0]
Available data
Summary of experiments/runs with jungfrau data available on 2018-04-19 is listed below.
Datasets with jungfrau data and detector Id
Dataset Runs DetInfo name Panel/detector Id or configuration object version ----------------------------------------------------------------------------------------------------------------------------- exp=mfx00616:run=8 1-8 MfxEndstation.0:Jungfrau.0 psana.Jungfrau.ConfigV1 exp=mfx05516:run=16 16 MfxEndstation.0:Jungfrau.0 171113-154920171025-3d00b0_171113-154920171025-3d00f7 exp=mfxlr1716:run=295 1-295 MfxEndstation.0:Jungfrau.0 171113-154920171025-3d00b0_171113-154920171025-3d00f7 exp=mfxlr5016:run=197 194-197 MfxEndstation.0:Jungfrau.0 171113-154920171025-3d00b0_171113-154920171025-3d00f7 exp=mfxx20515:run=2 1-2 MfxEndstation.0:Jungfrau.0 170505-149520170815-3d00b0_170505-149520170815-3d00f7 exp=mfx11116:run=697 563-697 MfxEndstation.0:Jungfrau.1 171113-154920171025-3d00fb exp=mfxls0816:run=193 1-193 MfxEndstation.0:Jungfrau.1 171113-154920171025-3d00fb exp=xcs11116:run=18 4-18 XcsEndstation.0:Jungfrau.0 psana.Jungfrau.ConfigV2 exp=xcsx22015:run=555 503-555 XcsEndstation.0:Jungfrau.0 psana.Jungfrau.ConfigV2 exp=xcsls3716:run=631 1-631 XcsEndstation.0:Jungfrau.0 170505-149520170815-3d00b0_170505-149520170815-3d00f7 exp=xcslr6316:run=17 4-17 XcsEndstation.0:Jungfrau.1 171113-154920171025-3d00fb exp=cxi11216:run=54 4-54 CxiEndstation.0:Jungfrau.0 psana.Jungfrau.ConfigV1 exp=detdaq17:run=68 3-68 DetLab.0:Jungfrau.0 171113-154920171025-3d00fb exp=xppls2416:run=247 1-247 XppEndstation.0:Jungfrau.1 171113-154920171025-3d00b0_171113-154920171025-3d00f7 exp=mecls3216:run=226 2-226 MecTargetChamber.0:Jungfrau.0 170505-149520170815-3d00b0_170505-149520170815-3d00f7
Panel calibration arrays
For now, calibration arrays for offset and gain is produced per panel named as M044, M068, M088, etc., e.g. directory /reg/g/psdm/detector/gains/jungfrau/M088/ contains:
g0_gain.npy g0_offset.npy g1_gain.npy g1_offset.npy g2_gain.npy g2_offset.npy
Content of calibration directory for panels
For auto-generation of calibration files it is suggested to
- keep panel calibration files under /reg/g/psdm/detector/gains/jungfrau/M044, M068, M088, etc.
- use panel Id with timestamp (version) as a reference to one of above directories
- in M044, M068, M088, etc. keep the same naming conventions for gain range and type (g1_gain.npy, g1_offset.npy, g2_gain.npy, etc.).
To distinguish calibration versions the time-stamp in format YYYYMMDDHHMMSS is dash-concatenated at the end of the panel Id name. This time stamp stands for start time of the constants validity range.
Assignment of panel ids to directories with calibration files can be listed by the command:
ls -l /reg/g/psdm/detector/gains/jungfrau
lrwxrwxrwx 1 dubrovin ps-data 4 Apr 30 11:56 170505-149520170815-3d00b0-20171025000000 -> M068 lrwxrwxrwx 1 dubrovin ps-data 4 Apr 30 11:56 170505-149520170815-3d00f7-20171025000000 -> M088 drwxr-sr-x 2 dubrovin ps-data 2 Apr 30 12:01 171113-154920171025-3d00b0 -> M068 drwxr-sr-x 2 dubrovin ps-data 2 Apr 30 12:01 171113-154920171025-3d00f7 -> M088 lrwxrwxrwx 1 dubrovin ps-data 4 Apr 30 11:55 171113-154920171025-3d00fb-20180306000000 -> M044 drwxr-sr-x 2 dubrovin ps-data 12 Mar 27 12:11 M044 drwxr-sr-x 2 dubrovin ps-data 11 Mar 27 12:10 M068 drwxr-sr-x 2 dubrovin ps-data 11 Mar 27 12:10 M088
At these conventions calibration file for particular detector can be generated dynamically using panel Ids from configuration data. Thus generated file can be deployed in the calib or "generic" repository.
CLI for calibration
Two commands are available in releases >ana-1.3.51
Command line to get jungfrau Id:
jungfrau_id <=== prints help jungfrau_id exp=xpptut15:run=430 [Jungfrau.0]
This command dumps full name of the detector found in configuration object, number of panels and detector Id:
Command line to generate and deploy gain correction constants (pixel_gain and pixel_offset) for specified experiment and detector:
jungfrau_gain_constants [-h] <=== prints help jungfrau_gain_constants -d exp=xcsls3716:run=631 -D [-s Jungfrau.0] <=== generate and deploy constants
This command
- retrieves detector Id, defines panel Ids
- retrieves gain correction files for panel Ids from specialized repository
/reg/g/psdm/detector/gains/jungfrau - merge them in arrays shaped as
(3,<number-of-pannels>,512,1024)for entire detector - saves array in files in local directory with names like
xcsls3716-r0631_XcsEndstation.0:Jungfrau.0_170505-149520170815-3d00b0_170505-149520170815-3d00f7_pixel_gain.txt - and deploy them under
/reg/d/psdm/XCS/xcsls3716/calib/Jungfrau::CalibV1/XcsEndstation.0:Jungfrau.0for calibration typespixel_gainpixel_offsetwith name631-end.datafor validity range beginning from run specified in dataset description.
Maintenance of calibration constants
Panel-calibration files
It is assumed that panel calibration constants for pixel_gain and pixel_offset are supplied by the detector group for all known panels. These constants should be available under directory /reg/g/psdm/detector/gains/jungfrau/ in files like
/reg/g/psdm/detector/gains/jungfrau/<panel-id>[-<time-stamp>]/g<index-of-gain-range>-<gain-or-offset>.npy.
Generator of constants
Example of the command to produce calibration constants for entire detector:
jungfrau_gain_constants -d exp=xcsls3716:run=631 [-D] [-U] [-s Jungfrau.0]
- constants are saved in local directory in files with long name containing fields of experiment, run, source, detector Id, constant type and extension txt or npy,
- option -d specifies dataset containing jungfrau data and configuration to define detector numeric Id, e.g. 170505-149520170815-3d00b0_170505-149520170815-3d00f7,
- option
-Ddeploys calibration constants in the calib directory for specified experiment and beginning of the validity range defined by the specified run number, - option
-Ushows commands (dcs add ...) to upload constants from local directory to global repository under/reg/d/psdm/detector/calib/jungfrau, - option
-sis required to distinguish between multiple detectors if specified data set has more than one jungfrau detector.
Upload constants to repository
Use command jungfrau_gain_constants with option -U to see commands for deployment (dcs add ...) . Then, apply these commands to add calibration constants to the detector associated hdf5 file located under directory
/reg/d/psdm/detector/calib/jungfrau//reg/g/psdm/detector/calib/jungfrau/- test repository
Questions
- format of integers in the panel name fields: hex(v).lsplit('0x'), e.g. "171113-154920171025-3d00fb"
- calibration versions: use time-stamp in format YYYYMMDDHHMMSS dash-concatenated to the panel Id, e.g. "171113-154920171025-3d00fb-2018030600000"
References
Overview
Content Tools