pbelange / cals2mask

Tools to create a config file at a given timestamp from NXCALS

CALS2Mask

About

The main tools of this package are found in Backend/Parser.py and Backend/Spark.py. To generate a config file to be used with lhcmask, one can modify the snapshot_ts in /make_config.py and run the script. config files for both B1 and B2 will be generated for the chosen timestamp.

The template used for the config file is found in Backend/config_template.yaml. The values to be replaced have been manually initialized to 000, their NXCALS variable name is stored in Backend/Parser -> Parser().varlist and their (nested) location in the config file is found in Backend/Parser -> Parser().configloc

Usage

Parser

With a DataFrame of data from NXCALS, the properties required by the config file from lhcmask can be extracted using the parsing module:

# linking DataFrame to Parser object
mainParser = parser.Parser(data_df)

# setting the observation timestamp:
mainParser.obs_timestamp('2022-07-23 19:30')

# setting the observation beam
mainParser.obs_beam('B1')

Then all the relevant quantities can be readily extracted from the DataFrame individually:

# Reading the attribute:
mainParser.beam_energy_tot

# Or equivalently, accessing the attribute:
mainParser['beam_energy_tot']

Or all at once:

>>> print(40*'-')
    for varName in mainParser.configloc.keys():
        print(f'{varName.ljust(18)}\t{mainParser[varName]}')
    print(40*'-')
    
----------------------------------------
mode              	b4_from_b2_with_bb
optics_file       	acc-models-lhc/operation/optics/R2022a_A30cmC30cmA10mL200cm.madx
beam_norm_emit_x  	1.4367134905573662
beam_norm_emit_y  	1.359021165398161
beam_sigt         	0.30069127446113253
beam_npart        	85156221602825
beam_energy_tot   	6799.6796875
nco_IP1           	974.0
nco_IP5           	974.0
nco_IP2           	876.0
nco_IP8           	912.0
on_x1             	-160.0
on_x2v            	200.0
on_x5             	160.0
on_x8h            	-200.0
----------------------------------------

All the required NXCALS variables can be listed to facilitate the extraction using the Spark module, using:

allvars = mainParser.get_varList()

Spark

To extract a DataFrame of data from NXCALS, one can use:

snapshot_ts = '2022-07-23 19:30'

# Initializing the spark session
sparkspace  = spark.SparkSession()

# Choosing a 24h window around the chosen ts
_start = str(pd.Timestamp(snapshot_ts)-pd.Timedelta(seconds = 24*3600))
_stop  = str(pd.Timestamp(snapshot_ts)+pd.Timedelta(seconds = 24*3600))

# Extracting data
var   = "HX:FILLN"
df    = sparkspace.query(var,_start,_stop)

Installation:

This package is intended to be used on a acc computer. We first need to install nxcals and the working python environment. Modify the last line of installme.sh with your username: ... [email protected]: ... and then:

bash installme.sh 

Then

source py_env/bin/activate 

To launch the spark builder, a kinit is needed:

kinit username

Warning : to ensure compatibility of xtrack with python 3.7 used in acc-py, one needs to change the 4 f-strings in ./py_env/xtrack/xtrack/loss_location_refinement as such:

# From 
f'{i_aper_1=}'
# To
f'i_aper_1={i_aper_1}'

This should be done automatically at the end of the installme.sh script.