def __init__(self,
files_in,
lh5_group,
dsp_config=None,
database=None,
n_drawn=1,
x_unit='ns',
x_lim=None,
waveforms='waveform',
wf_styles=None,
lines=None,
legend=None,
legend_opts=None,
norm=None,
align=None,
selection=None,
buffer_len=128,
block_width=8,
verbosity=1):
"""Constructor for WaveformBrowser:
- file_in: name of file or list of names to browse. Can use wildcards
- lh5_group: name of LH5 group in file to browse
- dsp_config (optional): name of DSP config json file containing transforms available to draw
- database (optional): dict with database of processing parameters
- n_drawn (default 1): number of events to draw simultaneously when calling DrawNext
- x_unit (default ns): unit for x-axis
- x_lim (default auto): range of x-values passes as tuple
- waveforms (default 'waveform'): name of wf or list of wf names to draw
- wf_styles (default None): waveform colors and other style parameters to cycle through when drawing waveforms. Can be given as:
dict of lists: e.g. {'color':['r', 'g', 'b'], 'linestyle':['-', '--', '.']}
name of predefined style; see matplotlib.style documentation
None: use current matplotlib style
If a single style cycle is given, use for all lines; if a list is given, match to waveforms list.
- lines (default None): name of parameter or list of parameters to draw hlines and vlines for
- legend (default None): formatting string and values to include in the legend. This can be a list of values (one for each waveform in waveforms). The values can be given as a tuple whose first entry is a formatting string and subsequent entries are the values to place in the formatting string. When building a formatting string, if a name is given in the {}s, it is assumed to be a parameter from the DSP config file. An example is:
("{:0.1f} keV", energy)
- legend_opts (default None): dict containing kwargs for formatting the legend
- norm (default None): name of parameter (probably energy) to use to normalize WFs; useful when drawing multiple
- align (default None): name of time parameter to set as 0 time; useful for aligning multiple waveforms
- selection (optional): selection of events to draw. Can be either a list of event indices or a numpy array mask (ala pandas).
- buffer_len (default 128): number of waveforms to keep in memory at a time
- block_width (default 8): block width for processing chain
"""
self.verbosity = verbosity
# data i/o initialization
self.lh5_st = lh5.Store(keep_open=True)
if isinstance(files_in, str): files_in = [files_in]
# Expand wildcards and map out the files
self.lh5_files = [
f for f_wc in files_in
for f in sorted(glob.glob(os.path.expandvars(f_wc)))
]
self.lh5_group = lh5_group
# file map is cumulative lenght of files up to file n. By doing searchsorted left, we can get the file for a given wf index
self.file_map = np.array(
[self.lh5_st.read_n_rows(lh5_group, f) for f in self.lh5_files],
'int64')
np.cumsum(self.file_map, out=self.file_map)
# Get the input buffer and read the first chunk
self.lh5_in = self.lh5_st.get_buffer(self.lh5_group, self.lh5_files[0],
buffer_len)
self.lh5_st.read_object(self.lh5_group,
self.lh5_files[0],
start_row=0,
n_rows=buffer_len,
obj_buf=self.lh5_in)
self.buffer_len = buffer_len
self.current_file = None
self.current_chunk = None
# initialize stuff for iteration
self.selection = selection
self.index_it = None
self.reset()
self.n_drawn = n_drawn
# initialize list of objects to draw
if isinstance(waveforms, str): self.wf_names = [waveforms]
elif waveforms is None: self.wf_names = []
else: self.wf_names = list(waveforms)
self.wf_data = [[] for _ in self.wf_names]
# wf_styles
if isinstance(wf_styles, list) or isinstance(wf_styles, tuple):
self.wf_styles = [None for _ in self.wf_data]
for i, sty in enumerate(wf_styles):
if isinstance(sty, str):
try:
self.wf_styles[i] = plt.style.library[sty][
'axes.prop_cycle']
except:
self.wf_styles[i] = itertools.repeat(None)
elif sty is None:
self.wf_styles[i] = itertools.repeat(None)
else:
self.wf_styles[i] = cycler(**sty)
else:
if isinstance(wf_styles, str):
try:
self.wf_styles = plt.style.library[wf_styles][
'axes.prop_cycle']
except:
self.wf_styles = itertools.repeat(None)
elif wf_styles is None:
self.wf_styles = itertools.repeat(None)
else:
self.wf_styles = cycler(**wf_styles)
if lines is None: self.line_names = []
elif isinstance(lines, list): self.line_names = lines
elif isinstance(lines, tuple): self.line_names = list(lines)
else: self.line_names = [lines]
self.line_data = [[] for _ in self.line_names]
if legend is None: legend = []
elif not isinstance(legend, list): legend = [legend]
# Set up the legend format strings and collect input values
self.legend_input = []
self.legend_format = []
for entry in legend:
legend_input = []
legend_format = ''
if not isinstance(entry, tuple):
entry = (entry, )
for val in entry:
if isinstance(val, str):
for st, name, form, cv in string.Formatter().parse(val):
legend_format += st
if name is not None:
legend_format += '{'
legend_input.append(name)
if form is not None and form != '':
legend_format += ':' + form
if cv is not None and cv != '':
legend_format += '!' + cv
legend_format += '}'
else:
# find any {}s to fill from the formatter
idxs = [
i for i, inp in enumerate(legend_input)
if isinstance(inp, str) and inp == ''
]
if idxs: # if we found a {}. it's already in the formatter
legend_input[idxs[0]] = val
else: # otherwise add to formatter
legend_input.append(val)
if legend_format != '': legend_format += ', '
if isinstance(val, pd.Series):
legend_format += val.name + ' = {:.3g}'
elif isinstance(val, np.ndarray):
legend_format += '{:.3g}'
self.legend_input.append(legend_input)
self.legend_format.append(legend_format)
self.legend_data = [[] for _ in self.legend_input]
self.legend_kwargs = legend_opts if legend_opts else {}
self.norm_par = norm
self.align_par = align
self.x_unit = units.unit_parser.parse_unit(x_unit)
self.x_lim = x_lim
# make processing chain and output buffer
outputs = self.wf_names + \
[name for name in self.line_names if isinstance(name, str)] + \
[name for name in self.legend_input if isinstance(name, str)]
if isinstance(self.norm_par, str): outputs += [self.norm_par]
if isinstance(self.align_par, str): outputs += [self.align_par]
self.proc_chain, self.lh5_out = build_processing_chain(
self.lh5_in,
dsp_config,
db_dict=database,
outputs=outputs,
verbosity=self.verbosity,
block_width=block_width)
self.fig = None
self.ax = None
def run_dsp(dfrow):
"""
run dsp on the test file, editing the processor list
alternate idea: generate a long list of processors with different names
"""
# adjust dsp config dictionary
rise, flat = dfrow
# dsp_config['processors']['wf_pz']['defaults']['db.pz.tau'] = f'{tau}*us'
dsp_config['processors']['wf_trap']['args'][1] = f'{rise}*us'
dsp_config['processors']['wf_trap']['args'][2] = f'{flat}*us'
# pprint(dsp_config)
# run dsp
pc, tb_out = build_processing_chain(tb_data, dsp_config, verbosity=0)
pc.execute()
# analyze peak
e_peak = 1460.
etype = 'trapEmax'
elo, ehi, epb = 4000, 4500, 3 # the peak moves around a bunch
energy = tb_out[etype].nda
# get histogram
hE, bins, vE = pgh.get_hist(energy, range=(elo, ehi), dx=epb)
xE = bins[1:]
# should I center the max at 1460?
# simple numerical width
i_max = np.argmax(hE)
h_max = hE[i_max]
upr_half = xE[(xE > xE[i_max]) & (hE <= h_max / 2)][0]
bot_half = xE[(xE < xE[i_max]) & (hE >= h_max / 2)][0]
fwhm = upr_half - bot_half
sig = fwhm / 2.355
# fit to gaussian: amp, mu, sig, bkg
fit_func = pgf.gauss_bkg
amp = h_max * fwhm
bg0 = np.mean(hE[:20])
x0 = [amp, xE[i_max], sig, bg0]
xF, xF_cov = pgf.fit_hist(fit_func, hE, bins, var=vE, guess=x0)
# collect results
e_fit = xF[0]
xF_err = np.sqrt(np.diag(xF_cov))
e_err = xF
fwhm_fit = xF[1] * 2.355 * 1460. / e_fit
fwhm_err = xF_err[2] * 2.355 * 1460. / e_fit
chisq = []
for i, h in enumerate(hE):
model = fit_func(xE[i], *xF)
diff = (model - h)**2 / model
chisq.append(abs(diff))
rchisq = sum(np.array(chisq) / len(hE))
fwhm_ovr_mean = fwhm_fit / e_fit
if show_movie:
plt.plot(xE,
hE,
ds='steps',
c='b',
lw=2,
label=f'{etype} {rise}--{flat}')
# peak shape
plt.plot(xE,
fit_func(xE, *x0),
'-',
c='orange',
alpha=0.5,
label='init. guess')
plt.plot(xE,
fit_func(xE, *xF),
'-r',
alpha=0.8,
label='peakshape fit')
plt.plot(np.nan,
np.nan,
'-w',
label=f'mu={e_fit:.1f}, fwhm={fwhm_fit:.2f}')
plt.xlabel(etype, ha='right', x=1)
plt.ylabel('Counts', ha='right', y=1)
plt.legend(loc=2)
# show a little movie
plt.show(block=False)
plt.pause(0.01)
plt.cla()
# return results
return pd.Series({
'e_fit': e_fit,
'fwhm_fit': fwhm_fit,
'rchisq': rchisq,
'fwhm_err': xF_err[0],
'fwhm_ovr_mean': fwhm_ovr_mean
})
def optimize_dcr(dg):
"""
I don't have an a priori figure of merit for the DCR parameter, until I can
verify that we're seeing alphas. So this function should just run processing
on a CAGE run with known alpha events, and show you the 2d DCR vs. energy.
Once we know we can reliably measure the alpha distribution somehow, then
perhaps we can try a grid search optimization like the one done in
optimize_trap.
"""
# files to consider. fixme: right now only works with one file
sto = lh5.Store()
lh5_dir = os.path.expandvars(dg.config['lh5_dir'])
raw_list = lh5_dir + dg.fileDB['raw_path'] + '/' + dg.fileDB['raw_file']
f_raw = raw_list.values[0]
tb_raw = 'ORSIS3302DecoderForEnergy/raw/'
tb_data = sto.read_object(tb_raw, f_raw)
cycle = dg.fileDB['cycle'].values[0]
f_results = f'./temp_{cycle}.h5'
write_output = True
# adjust dsp config
with open('opt_dcr.json') as f:
dsp_config = json.load(f, object_pairs_hook=OrderedDict)
# pprint(dsp_config)
# exit()
# set dcr parameters
# rise, flat, dcr_tstart = 200, 1000, 'tp_0+1.5*us' # default
# dcr_rise, dcr_flat, dcr_tstart = 100, 3000, 'tp_0+3*us' # best so far?
dcr_rise, dcr_flat, dcr_tstart = 100, 2500, 'tp_0+1*us'
dsp_config['processors']['dcr_raw']['args'][1] = dcr_rise
dsp_config['processors']['dcr_raw']['args'][2] = dcr_flat
dsp_config['processors']['dcr_raw']['args'][3] = dcr_tstart
# set trap energy parameters
# ene_rise, ene_flat = "2*us", "1*us" # best? from optimize_trap
ene_rise, ene_flat = "10*us", "5*us"
dsp_config['processors']['wf_trap']['args'][1] = ene_rise
dsp_config['processors']['wf_trap']['args'][2] = ene_flat
# adjust pole-zero constant
dsp_config['processors']['wf_pz']['defaults']['db.pz.tau'] = '64.4*us'
# dsp_config['processors']['wf_pz']['defaults']['db.pz.tau'] = '50*us'
# dsp_config['processors']['wf_pz']['defaults']['db.pz.tau'] = '100*us'
# run dsp
print('Running DSP ...')
t_start = time.time()
pc, tb_out = build_processing_chain(tb_data, dsp_config, verbosity=1)
pc.execute()
t_elap = (time.time() - t_start) / 60
print(f'Done. Elapsed: {t_elap:.2f} min')
df_out = tb_out.get_dataframe()
if write_output:
df_out.to_hdf(f_results, key='opt_dcr')
print('Wrote output file:', f_results)
def __init__(self,
files_in,
lh5_group,
dsp_config=None,
n_drawn=1,
x_unit='ns',
x_lim=None,
waveforms='waveform',
lines=None,
legend=None,
norm=None,
align=None,
selection=None,
buffer_len=128,
block_width=8,
verbosity=1):
"""Constructor for WaveformBrowser:
- file_in: name of file or list of names to browse. Can use wildcards
- lh5_group: name of LH5 group in file to browse
- dsp_config (optional): name of DSP config json file containing transforms available to draw
- n_drawn (default 1): number of events to draw simultaneously when calling DrawNext
- x_unit (default ns): unit for x-axis
- x_lim (default auto): range of x-values passes as tuple
- waveforms (default 'waveform'): name of wf or list of wf names to draw
- lines (default None): name of parameter or list of parameters to draw hlines and vlines for
- legend (default None): name of parameters to include in legend
- norm (default None): name of parameter (probably energy) to use to normalize WFs; useful when drawing multiple
- align (default None): name of time parameter to set as 0 time; useful for aligning multiple waveforms
- selection (optional): selection of events to draw. Can be either a list of event indices or a numpy array mask (ala pandas).
- buffer_len (default 128): number of waveforms to keep in memory at a time
- block_width (default 8): block width for processing chain
"""
self.verbosity = verbosity
# data i/o initialization
self.lh5_st = lh5.Store(keep_open=True)
if isinstance(files_in, str): files_in = [files_in]
# Expand wildcards and map out the files
self.lh5_files = [
f for f_wc in files_in
for f in sorted(glob.glob(os.path.expandvars(f_wc)))
]
self.lh5_group = lh5_group
# file map is cumulative lenght of files up to file n. By doing searchsorted left, we can get the file for a given wf index
self.file_map = np.array(
[self.lh5_st.read_n_rows(lh5_group, f) for f in self.lh5_files],
'int64')
np.cumsum(self.file_map, out=self.file_map)
# Get the input buffer and read the first chunk
self.lh5_in = self.lh5_st.get_buffer(self.lh5_group, self.lh5_files[0],
buffer_len)
self.lh5_st.read_object(self.lh5_group, self.lh5_files[0], 0,
buffer_len, self.lh5_in)
self.buffer_len = buffer_len
self.current_file = None
self.current_chunk = None
# initialize stuff for iteration
self.selection = selection
self.index_it = None
self.reset()
self.n_drawn = n_drawn
# initialize list of objects to draw
if isinstance(waveforms, str): self.waveforms = [waveforms]
elif waveforms is None: self.waveforms = []
else: self.waveforms = list(waveforms)
if isinstance(lines, str): self.lines = [lines]
elif lines is None: self.lines = []
else: self.lines = list(lines)
if isinstance(legend, str): self.legend = [legend]
elif legend is None: self.legend = []
else: self.legend = list(legend)
self.labels = []
self.norm_par = norm
self.align_par = align
self.x_unit = units.unit_parser.parse_unit(x_unit)
self.x_lim = x_lim
# make processing chain and output buffer
outputs = self.waveforms + self.lines + self.legend + (
[self.norm_par] if self.norm_par is not None else
[]) + ([self.align_par] if self.align_par is not None else [])
self.proc_chain, self.lh5_out = build_processing_chain(
self.lh5_in,
dsp_config,
outputs,
verbosity=self.verbosity,
block_width=block_width)
self.fig = None
self.ax = None
