class PseudoData[source]

PseudoData(func:Callable[int, ndarray], targ:Union[float, int])

Generic class for constructing pseudodata via sampling of random functions

d = PseudoData(lambda x: np.random.normal(0,1,(x,3)), 0)
df = d.get_df(10)
assert df.values.shape == (10,4)
df.head()

	0	1	2	gen_target
0	-0.059475	1.210967	-0.422888	0.0
1	-0.414044	1.268999	-0.580749	0.0
2	-0.697192	-0.487871	-1.681089	0.0
3	2.124146	1.264359	1.976336	0.0
4	2.111995	0.868064	2.019226	0.0

INFERNO paper

Callable class generating signal data. NB, paper states mu=[1,1], but plots and code indicate mu=[0,0]

s0 = paper_sig(2)
print(s0)
assert s0.shape == (2,3)

[[-1.44562602 -0.1492472   0.16018372]
 [-0.45275426 -0.71965287  0.08061889]]

paper_sig1 = _PaperData([0,0], [[1,0],[0,1]], 0, 2)
paper_sig1.r[0] += 1000
paper_sig1.r

array([1000,    0])

s1 = paper_sig1(2)
print(s1)
assert s1[0,0] > s0[0,0]

[[9.99702050e+02 1.78013375e+00 1.09574848e-01]
 [9.98718875e+02 1.58931519e+00 7.29373500e-01]]

Callable class generating background data with nuisances at nominal values

Check

n=1000
df = PseudoData(paper_sig, 1).get_df(n).append(PseudoData(paper_bkg, 0).get_df(n), ignore_index=True)
df

	0	1	2	gen_target
0	2.544023	0.442834	0.139470	1.0
1	-0.775732	0.719415	0.617672	1.0
2	-0.875915	-0.727330	0.406364	1.0
3	1.688095	0.688291	0.610886	1.0
4	0.528951	-0.639645	0.112357	1.0
...	...	...	...	...
1995	1.793329	1.821786	0.007961	0.0
1996	-1.923585	-1.982377	0.112923	0.0
1997	4.552509	-1.837093	0.467405	0.0
1998	-0.558329	0.875070	0.550241	0.0
1999	4.748507	-3.401783	0.527559	0.0

2000 rows × 4 columns

Reproduces paper plots

import seaborn as sns
sns.pairplot(df, hue='gen_target', vars=[f for f in df.columns if f != 'gen_target'])

/home/giles/anaconda3/lib/python3.6/site-packages/scipy/stats/stats.py:1713: FutureWarning: Using a non-tuple sequence for multidimensional indexing is deprecated; use `arr[tuple(seq)]` instead of `arr[seq]`. In the future this will be interpreted as an array index, `arr[np.array(seq)]`, which will result either in an error or a different result.
  return np.add.reduce(sorted[indexer] * weights, axis=axis) / sumval

<seaborn.axisgrid.PairGrid at 0x7f477dc44c18>