Getting Started with analysis_helpers¶

This notebook demonstrates two common operations:

• Asymmetry calculation from binned yields
• Efficiency and confidence intervals from selected/total samples

In [ ]:

import numpy as np
from analysis_helpers.asymmetry import get_asymmetry_array
from analysis_helpers.efficiency import get_efficiency

import numpy as np from analysis_helpers.asymmetry import get_asymmetry_array from analysis_helpers.efficiency import get_efficiency

1) Asymmetry from binned counts¶

In [ ]:

arr1 = np.array([120, 98, 105, 132, 111], dtype=float)
arr2 = np.array([115, 102, 95, 128, 120], dtype=float)

asym, asym_err = get_asymmetry_array(arr1, arr2)

print('asymmetry     =', np.round(asym, 4))
print('asymmetry err =', np.round(asym_err, 4))

arr1 = np.array([120, 98, 105, 132, 111], dtype=float) arr2 = np.array([115, 102, 95, 128, 120], dtype=float) asym, asym_err = get_asymmetry_array(arr1, arr2) print('asymmetry =', np.round(asym, 4)) print('asymmetry err =', np.round(asym_err, 4))

2) Efficiency with binned events¶

Here we create a toy dataset and emulate a simple selection threshold.

In [ ]:

rng = np.random.default_rng(42)

total = rng.normal(loc=0.0, scale=1.0, size=4000)
passed = total[total > -0.2]

bin_edges = np.linspace(-3.0, 3.0, 13)
eff, eff_cl = get_efficiency(passed, total, bin_edges)

print('efficiency =', np.round(eff, 3))
print('lower CL  =', np.round(eff_cl[0], 3))
print('upper CL  =', np.round(eff_cl[1], 3))

rng = np.random.default_rng(42) total = rng.normal(loc=0.0, scale=1.0, size=4000) passed = total[total > -0.2] bin_edges = np.linspace(-3.0, 3.0, 13) eff, eff_cl = get_efficiency(passed, total, bin_edges) print('efficiency =', np.round(eff, 3)) print('lower CL =', np.round(eff_cl[0], 3)) print('upper CL =', np.round(eff_cl[1], 3))

Next steps¶

• Replace toy arrays with your analysis variables
• Add event weights and pass them through the weighted API
• Visualize efficiency vs. the bin centers with matplotlib