JupyterHub

psana2 is also available in JupyterHub here in the kernel named "LCLS-II py3": https://pswww.slac.stanford.edu/jupyterhub/hub/home

Environment

To obtain the environment to run psana2, execute the following:

source /cds/sw/ds/ana/conda2/manage/bin/psconda.sh

Note that LCLS-II psana is not compatible with LCLS-I psana, so environments must activate one or the other, but not both.

Detector Names

Use this command to see non-epics detector names (see "Detector Interface" example below):

(ps-4.1.0) psanagpu101:lcls2$ detnames exp=tmoc00118,run=123
---------------------
Name      | Data Type
---------------------
epicsinfo | epicsinfo
timing    | raw      
hsd       | raw      
gmdstr0   | raw    
etc.

Use the same command with the "-e" option to see epics detector names (see "Detector Interface" example below). These are slowly varying variables (like temperatures) that are not strongly time correlated with the regular per-shot detector data:

(ps-4.1.0) psanagpu101:lcls2$ detnames -e exp=tmoc00118,run=123
---------------------------
Name            | Data Type
---------------------------
StaleFlags      | raw      
Keithley_Sum    | raw      
IM2K4_XrayPower | raw      
IM3K4_XrayPower | raw 
etc.

Using the Detector Interface

Standard (per-shot) detectors and the slower epics variables can be accessed as shown here using the names discovered with the commands above. You can use tab-completion in ipython or the jupyternotebook to explore what you can do with the various detector objects:

from psana import DataSource
ds = DataSource(exp='tmoc00118',run=123)
myrun = next(ds.runs())
opal = myrun.Detector('tmoopal')
epics_det = myrun.Detector('IM2K4_XrayPower')
for evt in myrun.events():
    img = opal.raw.image(evt)
    epics_val = epics_det(evt)
	# check for missing data
    if img is None or epics_val is None:
        print('none')
        continue
    print(img.shape,epics_val)

Example Script Producing Small HDF5 File

You can run this script with MPI: PS_SRV_NODES=2; mpirun -n 6 python example.py

It also works on one core with: python example.py. See MPI rank/task diagram here.

This mechanism by defaults produces "aligned" datasets where missing values are padded (with NaN's for floats, and -99999 for integers). To create an unaligned dataset (without padding) prefix the name of the variable with "unaligned_".

from psana import DataSource
import numpy as np

# called back on each SRV node, for every smd.event() call below
def test_callback(data_dict):
    print(data_dict)
ds = DataSource(exp='tmoc00118', run=123)
# batch_size here specifies how often the dictionary of information
# is sent to the SRV nodes
smd = ds.smalldata(filename='my.h5', batch_size=5, callbacks=[test_callback])
run = next(ds.runs())

# necessary (instead of "None") since some ranks may not receive events
# and the smd.sum() below could fail
arrsum = np.zeros((2), dtype=np.int)
for i,evt in enumerate(run.events()):
    myones = np.ones_like(arrsum)
    smd.event(evt, myfloat=2.0, arrint=myones)
    arrsum += myones

if smd.summary:
    smd.sum(arrsum)
    smd.save_summary({'summary_array' : arrsum}, summary_int=1)
smd.done()

MPI Task Structure

To allow for scaling, many hdf5 files are written, one per "SRV" node. The total number of SRV nodes is defined by the environment variable PS_SRV_NODES (defaults to 0). These many hdf5 files are joined by psana into what appears to be one file using the hdf5 "virtual dataset" feature.

Page tree

psana