def _remove_outliers_from_hist(
hist: Hist, outliers_start_index: int,
outliers_removal_axis: OutliersRemovalAxis) -> None:
"""Remove outliers from a given histogram.
Args:
hist: Histogram to check for outliers.
outliers_start_index: Index in the truth axis where outliers begin.
outliers_removal_axis: Axis along which outliers removal will be performed. Usually
the particle level aixs.
Returns:
None. The histogram is modified in place.
"""
# Use on TH1, TH2, and TH3 since we don't start removing immediately, but instead only after the limit
if outliers_start_index > 0:
# logger.debug("Removing outliers")
# Check for values above which they should be removed by translating the global index
x = ctypes.c_int(0)
y = ctypes.c_int(0)
z = ctypes.c_int(0)
# Maps axis to valaues
# This is kind of dumb, but it works.
outliers_removal_axis_values: Dict[OutliersRemovalAxis,
ctypes.c_int] = {
projectors.TH1AxisType.x_axis:
x,
projectors.TH1AxisType.y_axis:
y,
projectors.TH1AxisType.z_axis:
z,
}
for index in range(0, hist.GetNcells()):
# Get the bin x, y, z from the global bin
hist.GetBinXYZ(index, x, y, z)
# Watch out for any problems
if hist.GetBinContent(index) < hist.GetBinError(index):
logger.warning(
f"Bin content < error. Name: {hist.GetName()}, Bin content: {hist.GetBinContent(index)}, Bin error: {hist.GetBinError(index)}, index: {index}, ({x.value}, {y.value})"
)
if outliers_removal_axis_values[
outliers_removal_axis].value >= outliers_start_index:
# logger.debug("Cutting for index {}. x bin {}. Cut index: {}".format(index, x, cutIndex))
hist.SetBinContent(index, 0)
hist.SetBinError(index, 0)
else:
logger.info(f"Hist {hist.GetName()} did not have any outliers to cut")
def _from_th1(
hist: Hist
) -> Tuple[np.ndarray, np.ndarray, np.ndarray, Dict[str, Any]]:
""" Convert a TH1 histogram to a Histogram.
Note:
Underflow and overflow bins are excluded!
Args:
hist (ROOT.TH1): Input histogram.
Returns:
tuple: (x, y, errors) where x is the bin centers, y is the bin values, and
errors are the sumw2 bin errors.
"""
# Enable sumw2 if it's not already calculated
if hist.GetSumw2N() == 0:
hist.Sumw2(True)
# Don't include overflow
bin_edges = get_bin_edges_from_axis(hist.GetXaxis())
# NOTE: The y value and bin error are stored with the hist, not the axis.
y = np.array([
hist.GetBinContent(i)
for i in range(1,
hist.GetXaxis().GetNbins() + 1)
])
errors = np.array(hist.GetSumw2())
# Exclude the under/overflow bins
errors = errors[1:-1]
metadata = {}
# Check for a TProfile.
# In that case we need to retrieve the errors manually because the Sumw2() errors are
# not the anticipated errors.
if hasattr(hist, "BuildOptions"):
errors = np.array([
hist.GetBinError(i)
for i in range(1,
hist.GetXaxis().GetNbins() + 1)
])
# We expected errors squared
errors = errors**2
else:
# Sanity check. If they don't match, something odd has almost certainly occurred.
if not np.isclose(errors[0], hist.GetBinError(1)**2):
raise ValueError(
"Sumw2 errors don't seem to represent bin errors!")
# Retrieve the stats and store them in the metadata.
# They are useful for calculating histogram properties (mean, variance, etc).
stats = np.array([0, 0, 0, 0], dtype=np.float64)
hist.GetStats(np.ctypeslib.as_ctypes(stats))
# Return values are (each one is a single float):
# [1], [2], [3], [4]
# [1]: total_sum_w: Sum of weights (equal to np.sum(y) if unscaled)
# [2]: total_sum_w2: Sum of weights squared (equal to np.sum(errors_squared) if unscaled)
# [3]: total_sum_wx: Sum of w*x
# [4}: total_sum_wx2: Sum of w*x*x
metadata.update(_create_stats_dict_from_values(*stats))
return (bin_edges, y, errors, metadata)