ChangeLog.md 9.35 KB
Newer Older
1
2
3
Changes since NIFTy 6
=====================

4
5
6
7
CorrelatedFieldMaker interface change
-------------------------------------

The interface of `ift.CorrelatedFieldMaker.make` and
Martin Reinecke's avatar
Martin Reinecke committed
8
`ift.CorrelatedFieldMaker.add_fluctuations` changed; it now expects the mean
9
and the standard deviation of their various parameters not as separate
Martin Reinecke's avatar
Martin Reinecke committed
10
arguments but as a tuple.
11

12
Furthermore, it is now possible to disable the asperity and the flexibility
Martin Reinecke's avatar
Martin Reinecke committed
13
14
together with the asperity in the correlated field model. Note that disabling
only the flexibility is not possible.
15

16
17
18
SimpleCorrelatedField
---------------------

Martin Reinecke's avatar
Martin Reinecke committed
19
A simplified version of the correlated field model was introduced which does not
20
allow for multiple power spectra, the presence of a degree of freedom parameter
Martin Reinecke's avatar
Martin Reinecke committed
21
`dofdex`, or `total_N` larger than zero. Except for the above mentioned
22
23
limitations, it is equivalent to `ift.CorrelatedFieldMaker`. Hence, if one
wants to understand the implementation idea behind the model, it is easier to
Martin Reinecke's avatar
Martin Reinecke committed
24
grasp from reading `ift.SimpleCorrelatedField` than from going through
25
26
`ift.CorrelatedFieldMaker`.

Martin Reinecke's avatar
switch    
Martin Reinecke committed
27
28
29
30
31
32
Change in external dependencies
-------------------------------

Instead of the optional external packages `pypocketfft` and `pyHealpix`, NIFTy
now uses the DUCC package (<https://gitlab.mpcdf.mpg.de/mtr/ducc)>,
which is their successor.
33
34


Philipp Arras's avatar
Philipp Arras committed
35
36
37
38
39
40
Naming of operator tests
------------------------

The implementation tests for nonlinear operators are now available in
`ift.extra.check_operator()` and for linear operators
`ift.extra.check_linear_operator()`.
41

42

43
44
45
MetricGaussianKL interface
--------------------------

46
Users do not instantiate `MetricGaussianKL` by its constructor anymore. Rather
47
`MetricGaussianKL.make()` shall be used.
48
49
50
51


Changes since NIFTy 5
=====================
Martin Reinecke's avatar
Martin Reinecke committed
52

Martin Reinecke's avatar
Martin Reinecke committed
53
Minimum Python version increased to 3.6
54
---------------------------------------
Martin Reinecke's avatar
Martin Reinecke committed
55
56


57
New operators
58
-------------
59
60
61
62
63
64
65
66
67
68

In addition to the below changes, the following operators were introduced:

* UniformOperator: Transforms a Gaussian into a uniform distribution
* VariableCovarianceGaussianEnergy: Energy operator for inferring covariances
* MultiLinearEinsum: Multi-linear version of numpy's einsum with derivates
* LinearEinsum: Linear version of numpy's einsum with one free field
* PartialConjugate: Conjugates parts of a multi-field
* SliceOperator: Geometry preserving mask operator
* SplitOperator: Splits a single field into a multi-field
69
* MatrixProductOperator: Applies matrices (scipy.sparse, numpy) to fields
70
* IntegrationOperator: Integrates over subspaces of fields
71

Martin Reinecke's avatar
Martin Reinecke committed
72
FFT convention adjusted
73
-----------------------
Martin Reinecke's avatar
Martin Reinecke committed
74
75
76
77
78

When going to harmonic space, NIFTy's FFT operator now uses a minus sign in the
exponent (and, consequently, a plus sign on the adjoint transform). This
convention is consistent with almost all other numerical FFT libraries.

Martin Reinecke's avatar
Martin Reinecke committed
79
Interface change in EndomorphicOperator.draw_sample()
80
-----------------------------------------------------
Martin Reinecke's avatar
Martin Reinecke committed
81

Philipp Arras's avatar
Philipp Arras committed
82
Both complex-valued and real-valued Gaussian probability distributions have
Martin Reinecke's avatar
Martin Reinecke committed
83
Hermitian and positive endomorphisms as covariance. Just by looking at an
Philipp Arras's avatar
Philipp Arras committed
84
85
86
87
88
89
endomorphic operator itself it is not clear whether it is viewed as covariance
for real or complex Gaussians when a sample of the respective distribution shall
be drawn. Therefore, we introduce the method `draw_sample_with_dtype()` which
needs to be given the data type of the probability distribution. This function
is implemented for all operators which actually draw random numbers
(`DiagonalOperator` and `ScalingOperator`). The class `SamplingDtypeSetter` acts
Martin Reinecke's avatar
Martin Reinecke committed
90
as a wrapper for this kind of operators in order to fix the data type of the
Philipp Arras's avatar
Philipp Arras committed
91
92
93
94
95
distribution. Samples from these operators can be drawn with `.draw_sample()`.
In order to dive into those subtleties I suggest running the following code and
playing around with the dtypes.

```
Martin Reinecke's avatar
Martin Reinecke committed
96
import nifty7 as ift
Philipp Arras's avatar
Philipp Arras committed
97
98
99
100
101
import numpy as np

dom = ift.UnstructuredDomain(5)
dtype = [np.float64, np.complex128][1]
invcov = ift.ScalingOperator(dom, 3)
Martin Reinecke's avatar
Martin Reinecke committed
102
e = ift.GaussianEnergy(mean=ift.from_random(dom, 'normal', dtype=dtype),
Philipp Arras's avatar
Philipp Arras committed
103
                       inverse_covariance=invcov)
Martin Reinecke's avatar
Martin Reinecke committed
104
pos = ift.from_random(dom, 'normal', dtype=np.complex128)
Philipp Arras's avatar
Philipp Arras committed
105
106
107
108
109
lin = e(ift.Linearization.make_var(pos, want_metric=True))
met = lin.metric
print(met)
print(met.draw_sample())
```
Martin Reinecke's avatar
Martin Reinecke committed
110

111
112
113
114
115
116
New approach for sampling complex numbers
=========================================

When calling draw_sample_with_dtype with a complex dtype,
the variance is now used for the imaginary part and real part separately.
This is done in order to be consistent with the Hamiltonian.
117
Note that by this,
118
119
```
np.std(ift.from_random(domain, 'normal', dtype=np.complex128).val)
120
````
121
122
123
does not give 1, but sqrt(2) as a result.


Philipp Arras's avatar
Philipp Arras committed
124
MPI parallelisation over samples in MetricGaussianKL
125
----------------------------------------------------
Philipp Arras's avatar
Philipp Arras committed
126
127
128
129

The classes `MetricGaussianKL` and `MetricGaussianKL_MPI` have been unified
into one `MetricGaussianKL` class which has MPI support built in.

Martin Reinecke's avatar
Martin Reinecke committed
130
New approach for random number generation
131
-----------------------------------------
Martin Reinecke's avatar
Martin Reinecke committed
132
133
134
135
136

The code now uses `numpy`'s new `SeedSequence` and `Generator` classes for the
production of random numbers (introduced in numpy 1.17. This greatly simplifies
the generation of reproducible random numbers in the presence of MPI parallelism
and leads to cleaner code overall. Please see the documentation of
Martin Reinecke's avatar
Martin Reinecke committed
137
`nifty7.random` for details.
Martin Reinecke's avatar
Martin Reinecke committed
138

139

Rouven Lemmerz's avatar
Rouven Lemmerz committed
140
Interface Change for from_random and OuterProduct
141
-------------------------------------------------
Rouven Lemmerz's avatar
Rouven Lemmerz committed
142
143
144
145
146

The sugar.from_random, Field.from_random, MultiField.from_random now take domain
as the first argument and default to 'normal' for the second argument.
Likewise OuterProduct takes domain as the first argument and a field as the second.

Philipp Arras's avatar
Philipp Arras committed
147
Interface Change for non-linear Operators
148
-----------------------------------------
Philipp Arras's avatar
Philipp Arras committed
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163

The method `Operator.apply()` takes a `Linearization` or a `Field` or a
`MultiField` as input. This has not changed. However, now each non-linear
operator assumes that the input `Linearization` comes with an identity operator
as jacobian. Also it is assumed that the `apply()` method returns a
`Linearization` with the jacobian of the operator itself. The user is not in
charge anymore of stacking together the jacobians of operator chains. Something
like `x.jac` should not appear in any self-written `apply()` methods. The method
`Operator._check_input` tests if this condition is met. The same goes for the
metric. There is no need anymore to call `SandwichOperator` in an `apply()`
method when implementing new energies. This change should not lead to unexpected
behaviour since both `Operator._check_input()` and
`extra.check_jacobian_consistency()` tests for the new conditions to be
fulfilled.

Martin Reinecke's avatar
Martin Reinecke committed
164
Special functions for complete Field reduction operations
165
---------------------------------------------------------
Martin Reinecke's avatar
Martin Reinecke committed
166
167
168
169
170
171
172
173
174
175

So far, reduction operations called on Fields (like `vdot`, `sum`, `integrate`,
`mean`, `var`, `std`, `prod` etc.) returned a scalar when the reduction was
carried out over all domains, and otherwise a `Field`.
Having the data type of the returned value depend on input parameters is
extremely confusing, so all of these reduction operations now always return a
Field. We also introduced another set of reduction operations which always
operate over all subdomains and therefore don't take a `spaces` argument; they
are named `s_vdot`, `s_sum` etc. and always return a scalar.

176
Updates regarding correlated fields
177
-----------------------------------
178
179
180
181
182
183
184
185
186
187
188
189
190
191

The most commonly used model for homogeneous and isotropic correlated fields in
nifty5 has been `SLAmplitude` combined with `CorrelatedField`. This model
exhibits unintuitive couplings between its parameters and as been replaced
by `CorrelatedFieldMaker` in NIFTy 6. This model aims to conceptionally provide
the same functionality. However, internally it works quite differently. Therefore,
specific classes for `SLAmplitude` like `LogRGSpace`, `QHTOperator`, `ExpTransform`,
`SlopeOperator`, `SymmetrizingOperator`, `CepstrumOperator`, `CorrelatedField`
and `MfCorrelatedField` are not needed anymore and have been removed. In general,
`CorrelatedFieldMaker` feels to be better conditioned leading to faster convergence
but it is hard to make explicit tests since the two approaches cannot be mapped
onto each other exactly. We experienced that preconditioning in the `MetricGaussianKL`
via `napprox` breaks the inference scheme with the new model so `napprox` may not
be used here.
Martin Reinecke's avatar
merge    
Martin Reinecke committed
192

Lukas Platz's avatar
Lukas Platz committed
193
Removal of the standard MPI parallelization scheme:
194
---------------------------------------------------
Martin Reinecke's avatar
Martin Reinecke committed
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211

When several MPI tasks are present, NIFTy5 distributes every Field over these
tasks by splitting it along the first axis. This approach to parallelism is not
very efficient, and it has not been used by anyone for several years, so we
decided to remove it, which led to many simplifications within NIFTy.
User-visible changes:
- the methods `to_global_data`, `from_global_data`, `from_local_data` and
  the property `local_data` have been removed from `Field` and `MultiField`.
  Instead there are now the property `val` (returning a read-only numpy.ndarray
  for `Field` and a dictionary of read-only numpy.ndarrays for `MultiField`) and
  the method `val_rw()` (returning the same structures with writable copies of
  the arrays). Fields and MultiFields can be created from such structures using
  the static method `from_raw`
- the functions `from_global_data` and `from_local_data` in `sugar` have been
  replaced by a single function called `makeField`
- the property `local_shape` has been removed from `Domain` (and subclasses)
  and `DomainTuple`.