1. Example scripts¶
This is an index of the example scripts in the examples directory in the
repository. The links point to the file preview on github.
1.1. Short index¶
a acic b bart barteasy c d dft doubleint e even f fourier g h i j k l m n o pdf1 pdf2 pdf3 pdf4 pdf5 pdf6 pdf7 pdf8 pdf9 q r s t u v w x y z
1.2. Long index¶
a.py: EXAMPLE A.
Where the oscillating nature of an unknown function is revealed from but a few points, though only to a certain distance.
acic.py: Analyze a dataset from the ACIC 2022 Data Challenge using the BART kernel.
Website: https://acic2022.mathematica.org
Article: Dan R.C. Thal and Mariel M. Finucane, “Causal Methods Madness: Lessons Learned from the 2022 ACIC Competition to Estimate Health Policy Impacts,” Observational Studies, Volume 9, Issue 3, 2023, pp. 3-27, https://doi.org/10.1353/obs.2023.0023
b.py: EXAMPLE B.
Where it is discovered that the derivative of the unknown function is orthogonal to the function itself, and furthermore that it is orange instead of blue.
bart.py: Example usage of the BART kernel to replace the standard BART MCMC algorithm.
barteasy.py: BART with the simplified subpackage.
c.py: EXAMPLE C.
Where a nonlinear transformation hides the true height of some crosses.
d.py: EXAMPLE D.
Where the primitive of our function takes some freedom to move up and down.
dft.py: Constrain the discrete Fourier transform of a periodic process. Shows how to use GP.addlintransf.
doubleint.py: Test of double integral constraint
e.py: EXAMPLE E.
Where observing both a function and its derivative put some restraint on their behaviour.
even.py: Split a function into even and odd parts.
f.py: EXAMPLE F.
Where apparently in these times it is not anymore possible to know exactly where one gentleman’s function will pass.
fourier.py: Constrain the values of Fourier series coefficients
g.py: EXAMPLE G.
Where two ways of expressing one’s beliefs are compared and found, satisfactorily, to be quite similar.
h.py: EXAMPLE H.
Where at first sight nothing has changed, but behind the scenes important information has been lost forever.
i.py: EXAMPLE I.
Where, due to obscure political reasons, we insist on forgetting important details that a reasonable man’s mind would be fond of recalling later.
j.py: EXAMPLE J.
Where the excessive smoothness of the prediction is found not to satisfy our manly tastes.
k.py: EXAMPLE K.
Where lady K finds out that every gentleman hides his rougher corners in his derivative.
l.py: EXAMPLE L.
Where two formulas give the same results and so math triumphs once again.
m.py: EXAMPLE M.
Where we discover that, unlike elephants, Matérn processes prefer to forget after less than one data step.
n.py: EXAMPLE N.
Where we wonder how much a derivative is allowed to do her own business compared to her mistress.
o.py: EXAMPLE O.
Where mister N’s enquiry is repeated by different means.
pdf1.py: Fit of parton distributions functions (PDFs)
pdf2.py: Fit of parton distributions functions (PDFs)
The difference from pdf1.py is that we define the transformation on processes instead of on their finite realizations
pdf3.py: Fit of parton distributions functions (PDFs)
Like pdf2, but with correct integral constraints and naming this time
pdf4.py: Fit of parton distribution functions (PDFs)
Like pdf3, but with hyperparameters
pdf5.py: Fit of parton distributions functions (PDFs)
Like pdf3, but with nonlinear data
pdf6.py: Fit of parton distributions functions (PDFs)
Like pdf5, but with uncertainties on M and M2
pdf7.py: Fit of parton distributions functions (PDFs)
Like pdf6, but with hyperparameters
pdf8.py: Fit of parton distributions functions (PDFs)
Like pdf7, but with more realistic PDFs
pdf9.py: Fit of parton distributions functions (PDFs)
Like pdf8, but only with linear data
q.py: EXAMPLE Q.
Where we extend a pattern of waves in space, but do not dare look too far from our data.
r.py: EXAMPLE R.
Where we decide to introduce a strong anisotropy despite evidence of its absence.
s.py: EXAMPLE S.
Where different coordinates unite together under a single field name.
t.py: EXAMPLE T.
Where we pretend to discover that two series of events were in fact one the delayed and imperfect copy of the other.
u.py: EXAMPLE U.
Where we infer the temporal scale of a process assuming another process is correlated with its derivative.
v.py: EXAMPLE V.
Where we go on an expedition to survey the many and wondrous kernels that inhabit our software.
w.py: EXAMPLE W.
Where, with limited success, we recover the identity of two mixed functions knowing their speed of variation.
x.py: EXAMPLE X.
Where the derivatives of an interesting correlation function are put to harsh a trial.
y.py: EXAMPLE Y.
Where a Zeta kernel forces some random samples to have zero mean.
z.py: EXAMPLE Z.
Where we sail in an infinite dimensional space to sum two numbers.
