Mercurial > repos > public > sbplib_julia

annotate Notes.md @ 829:1de10f5b2008 operator_storage_array_of_table

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
Add some notes on the format of operator storage
author Jonatan Werpers <jonatan@werpers.com>
date Wed, 12 Jan 2022 13:07:43 +0100
parents 1d166e7a9deb
children fe8fe3f01162
Ignore whitespace changes - Everywhere: Within whitespace: At end of lines:
rev   line source
318
0c8d4a734c4f Add some todos and add a Notes.md file
Jonatan Werpers <jonatan@werpers.com>
parents:
diff changeset
1 # Notes
0c8d4a734c4f Add some todos and add a Notes.md file
Jonatan Werpers <jonatan@werpers.com>
parents:
diff changeset
2
654
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
3 ## Reading operators
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
4
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
5 Jonatan's suggestion is to add methods to `Laplace`, `SecondDerivative` and
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
6 similar functions that take in a filename from which to read stencils. These
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
7 methods encode how to use the structure in a file to build the particular
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
8 operator. The filename should be a keyword argument and could have a default
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
9 value.
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
10
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
11 * This allows easy creation of operators without the user having to handle stencils.
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
12 * The user can easily switch between sets of operators by changing the file stecils are read from.
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
13
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
14 Grids for optimized operators could be created by reading from a .toml file in
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
15 a similar fashion as for the operators. The grid can then be used in a
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
16 `Laplace` method which dispatches on the grid type and knows how to read the
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
17 optimized operators. The method would also make sure the operators match the
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
18 grid.
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
19
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
20 Idea: Make the current upper case methods lower case. Add types with the upper
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
21 case names. These types are tensor mappings for the operator but also contain
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
22 the associated operators as fields. For example:
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
23
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
24 ```julia
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
25 L = Laplace(grid)
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
26 L.H
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
27 L.Hi
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
28 L.e
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
29 L.d
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
30 L.M
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
31
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
32 wave = L - L.Hi∘L.e'∘L.d
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
33 ```
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
34
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
35 These types could also contain things like borrowing and such.
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
36
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
37 ## Storage of operators
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
38 We need to change the toml format so that it is easier to store several
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
39 operator with different kinds of differentiations. For example there could be
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
40 several operators of the same order but with different number of boundary
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
41 points or different choice of boundary stencils.
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
42
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
43 Properties that differentiate operators should for this reason be stored in
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
44 variables and not be section or table names.
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
45
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
46 Operators/sets of stencils should be stored in an [array of tables](https://toml.io/en/v1.0.0-rc.3#array-of-tables).
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
47
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
48 We should formalize the format and write simple and general access methods for
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
49 getting operators/sets of stencils from the file. They should support a simple
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
50 way to filter based on values of variables. There filters could possibly be
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
51 implemented through keyword arguments that are sent through all the layers of
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
52 operator creation.
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
53
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
54 * Remove order as a table name and put it as a variable.
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
55
829
1de10f5b2008 Add some notes on the format of operator storage
Jonatan Werpers <jonatan@werpers.com>
parents: 775
diff changeset
56 ### Parsing
1de10f5b2008 Add some notes on the format of operator storage
Jonatan Werpers <jonatan@werpers.com>
parents: 775
diff changeset
57 At the moment the only parsing that can be done at the top level is conversion
1de10f5b2008 Add some notes on the format of operator storage
Jonatan Werpers <jonatan@werpers.com>
parents: 775
diff changeset
58 from the toml file to a dict of strings. This forces the user to dig through
1de10f5b2008 Add some notes on the format of operator storage
Jonatan Werpers <jonatan@werpers.com>
parents: 775
diff changeset
59 the dictionary and apply the correct parsing methods for the different parts,
1de10f5b2008 Add some notes on the format of operator storage
Jonatan Werpers <jonatan@werpers.com>
parents: 775
diff changeset
60 e.g. `parse_stencil` or `parse_tuple`. While very flexible there is a tight
1de10f5b2008 Add some notes on the format of operator storage
Jonatan Werpers <jonatan@werpers.com>
parents: 775
diff changeset
61 coupling between what is written in the file and what code is run to make data
1de10f5b2008 Add some notes on the format of operator storage
Jonatan Werpers <jonatan@werpers.com>
parents: 775
diff changeset
62 in the file usable. While this coupling is hard to avoid it should be made
1de10f5b2008 Add some notes on the format of operator storage
Jonatan Werpers <jonatan@werpers.com>
parents: 775
diff changeset
63 explicit. This could be done by putting a reference to a parsing function in
1de10f5b2008 Add some notes on the format of operator storage
Jonatan Werpers <jonatan@werpers.com>
parents: 775
diff changeset
64 the operator-storage format or somehow specifying the type of each object.
1de10f5b2008 Add some notes on the format of operator storage
Jonatan Werpers <jonatan@werpers.com>
parents: 775
diff changeset
65 This mechanism should be extensible without changing the package. Perhaps
1de10f5b2008 Add some notes on the format of operator storage
Jonatan Werpers <jonatan@werpers.com>
parents: 775
diff changeset
66 there could be a way to register parsing functions or object types for the
1de10f5b2008 Add some notes on the format of operator storage
Jonatan Werpers <jonatan@werpers.com>
parents: 775
diff changeset
67 toml.
1de10f5b2008 Add some notes on the format of operator storage
Jonatan Werpers <jonatan@werpers.com>
parents: 775
diff changeset
68
654
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
69
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
70 ## Variable second derivative
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
71
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
72 2020-12-08 after discussion with Vidar:
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
73 We will have to handle the variable second derivative in a new variant of
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
74 VolumeOperator, "SecondDerivativeVariable?". Somehow it needs to know about
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
75 the coefficients. They should be provided as an AbstractVector. Where they are
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
76 provided is another question. It could be that you provide a reference to the
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
77 array to the constructor of SecondDerivativeVariable. If that array is mutable
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
78 you are free to change it whenever and the changes should propagate
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
79 accordingly. Another option is that the counter part to "Laplace" for this
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
80 variable second derivate returns a function or acts like a functions that
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
81 takes an Abstract array and returns a SecondDerivativeVariable with the
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
82 appropriate array. This would allow syntax like `D2(a)*v`. Can this be made
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
83 performant?
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
84
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
85 For the 1d case we can have a constructor
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
86 `SecondDerivativeVariable(D2::SecondDerivativeVariable, a)` that just creates
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
87 a copy with a different `a`.
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
88
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
89 Apart from just the second derivative in 1D we need operators for higher
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
90 dimensions. What happens if a=a(x,y)? Maybe this can be solved orthogonally to
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
91 the `D2(a)*v` issue, meaning that if a constant nD version of
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
92 SecondDerivativeVariable is available then maybe it can be wrapped to support
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
93 function like syntax. We might have to implement `SecondDerivativeVariable`
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
94 for N dimensions which takes a N dimensional a. If this could be easily
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
95 closured to allow D(a) syntax we would have come a long way.
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
96
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
97 For `Laplace` which might use a variable D2 if it is on a curvilinear grid we
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
98 might want to choose how to calculate the metric coefficients. They could be
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
99 known on closed form, they could be calculated from the grid coordinates or
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
100 they could be provided as a vector. Which way you want to do it might change
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
101 depending on for example if you are memory bound or compute bound. This choice
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
102 cannot be done on the grid since the grid shouldn't care about the computer
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
103 architecture. The most sensible option seems to be to have an argument to the
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
104 `Laplace` function which controls how the coefficients are gotten from the
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
105 grid. The argument could for example be a function which is to be applied to
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
106 the grid.
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
107
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
108 What happens if the grid or the varible coefficient is dependent on time?
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
109 Maybe it becomes important to support `D(a)` or even `D(t,a)` syntax in a more
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
110 general way.
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
111
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
112 ```
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
113 g = TimeDependentGrid()
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
114 L = Laplace(g)
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
115 function Laplace(g::TimeDependentGrid)
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
116 g_logical = logical(g) # g_logical is time independent
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
117 ... Build a L(a) assuming we can do that ...
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
118 a(t) = metric_coeffs(g,t)
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
119 return t->L(a(t))
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
120 end
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
121 ```
d26231227b89 Add a bunch of notes on reading and storing operators and how to implement variable second derivatives
Jonatan Werpers <jonatan@werpers.com>
parents: 606
diff changeset
122
318
0c8d4a734c4f Add some todos and add a Notes.md file
Jonatan Werpers <jonatan@werpers.com>
parents:
diff changeset
123 ## Known size of range and domain?
320
d705b397aa33 Clean up Notes.md and TODO.md
Jonatan Werpers <jonatan@werpers.com>
parents: 319
diff changeset
124 Is there any reason to use a trait to differentiate between fixed size and unknown size?
d705b397aa33 Clean up Notes.md and TODO.md
Jonatan Werpers <jonatan@werpers.com>
parents: 319
diff changeset
125
379
de4746d6d126 Add some notes and todos
Jonatan Werpers <jonatan@werpers.com>
parents: 332
diff changeset
126 When do we need to know the size of the range and domain?
de4746d6d126 Add some notes and todos
Jonatan Werpers <jonatan@werpers.com>
parents: 332
diff changeset
127 * When indexing to provide boundschecking?
de4746d6d126 Add some notes and todos
Jonatan Werpers <jonatan@werpers.com>
parents: 332
diff changeset
128 * When doing specialised computations for different parts of the range/domain?
de4746d6d126 Add some notes and todos
Jonatan Werpers <jonatan@werpers.com>
parents: 332
diff changeset
129 * More?
319
05ac7d7f9176 Move research and thinking from TODO.md to Notes.md
Jonatan Werpers <jonatan@werpers.com>
parents: 318
diff changeset
130
379
de4746d6d126 Add some notes and todos
Jonatan Werpers <jonatan@werpers.com>
parents: 332
diff changeset
131 Maybe if we should have dynamic sizing it could be only for the range. `domain_size` would not be implemented. And the `range_size` would be a function of a vector that the TensorMapping is applied to.
332
535f1bff4bcc Add some notes and some todos
Jonatan Werpers <jonatan@werpers.com>
parents: 320
diff changeset
132
319
05ac7d7f9176 Move research and thinking from TODO.md to Notes.md
Jonatan Werpers <jonatan@werpers.com>
parents: 318
diff changeset
133 ## Reasearch and thinking
05ac7d7f9176 Move research and thinking from TODO.md to Notes.md
Jonatan Werpers <jonatan@werpers.com>
parents: 318
diff changeset
134 - [ ] Use a trait to indicate that a TensorMapping har the same range and domain?
05ac7d7f9176 Move research and thinking from TODO.md to Notes.md
Jonatan Werpers <jonatan@werpers.com>
parents: 318
diff changeset
135 - [ ] Rename all the Tensor stuff to just LazyOperator, LazyApplication and so on?
05ac7d7f9176 Move research and thinking from TODO.md to Notes.md
Jonatan Werpers <jonatan@werpers.com>
parents: 318
diff changeset
136 - [ ] Check how the native julia doc generator works
05ac7d7f9176 Move research and thinking from TODO.md to Notes.md
Jonatan Werpers <jonatan@werpers.com>
parents: 318
diff changeset
137 - [ ] Check if Vidars design docs fit in there
05ac7d7f9176 Move research and thinking from TODO.md to Notes.md
Jonatan Werpers <jonatan@werpers.com>
parents: 318
diff changeset
138 - [ ] Create a macro @lazy which replaces a binary op (+,-) by its lazy equivalent? Would be a neat way to indicate which evaluations are lazy without cluttering/confusing with special characters.
05ac7d7f9176 Move research and thinking from TODO.md to Notes.md
Jonatan Werpers <jonatan@werpers.com>
parents: 318
diff changeset
139 - [ ] Specificera operatorer i TOML eller något liknande?
05ac7d7f9176 Move research and thinking from TODO.md to Notes.md
Jonatan Werpers <jonatan@werpers.com>
parents: 318
diff changeset
140 H.. H_gamma etc.)
606
8f9b3eac128a Add a bunch of stuff to TODO.md and Notes.md
Jonatan Werpers <jonatan@werpers.com>
parents: 427
diff changeset
141 - [ ] Dispatch on Lower() instead of the type Lower so `::Lower` instead of `::Type{Lower}` ???
332
535f1bff4bcc Add some notes and some todos
Jonatan Werpers <jonatan@werpers.com>
parents: 320
diff changeset
142 Seems better unless there is some specific reason to use the type instead of the value.
379
de4746d6d126 Add some notes and todos
Jonatan Werpers <jonatan@werpers.com>
parents: 332
diff changeset
143 - [ ] How do we handle mixes of periodic and non-periodic grids? Seems it should be supported on the grid level and on the 1d operator level. Between there it should be transparent.
de4746d6d126 Add some notes and todos
Jonatan Werpers <jonatan@werpers.com>
parents: 332
diff changeset
144 - [ ] Can we have a trait to tell if a TensorMapping is transposable?
de4746d6d126 Add some notes and todos
Jonatan Werpers <jonatan@werpers.com>
parents: 332
diff changeset
145 - [ ] Is it ok to have "Constructors" for abstract types which create subtypes? For example a Grids() functions that gives different kind of grids based on input?
410
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
146
606
8f9b3eac128a Add a bunch of stuff to TODO.md and Notes.md
Jonatan Werpers <jonatan@werpers.com>
parents: 427
diff changeset
147 ## Regions and tensormappings
8f9b3eac128a Add a bunch of stuff to TODO.md and Notes.md
Jonatan Werpers <jonatan@werpers.com>
parents: 427
diff changeset
148 - [ ] Use a trait to indicate if a TensorMapping uses indices with regions.
8f9b3eac128a Add a bunch of stuff to TODO.md and Notes.md
Jonatan Werpers <jonatan@werpers.com>
parents: 427
diff changeset
149 The default should be that they do NOT.
8f9b3eac128a Add a bunch of stuff to TODO.md and Notes.md
Jonatan Werpers <jonatan@werpers.com>
parents: 427
diff changeset
150 - [ ] What to name this trait? Can we call it IndexStyle but not export it to avoid conflicts with Base.IndexStyle?
8f9b3eac128a Add a bunch of stuff to TODO.md and Notes.md
Jonatan Werpers <jonatan@werpers.com>
parents: 427
diff changeset
151 - [ ] Figure out repeated application of regioned TensorMappings. Maybe an instance of a tensor mapping needs to know the exact size of the range and domain for this to work?
8f9b3eac128a Add a bunch of stuff to TODO.md and Notes.md
Jonatan Werpers <jonatan@werpers.com>
parents: 427
diff changeset
152
8f9b3eac128a Add a bunch of stuff to TODO.md and Notes.md
Jonatan Werpers <jonatan@werpers.com>
parents: 427
diff changeset
153 ## Boundschecking and dimension checking
8f9b3eac128a Add a bunch of stuff to TODO.md and Notes.md
Jonatan Werpers <jonatan@werpers.com>
parents: 427
diff changeset
154 Does it make sense to have boundschecking only in getindex methods?
8f9b3eac128a Add a bunch of stuff to TODO.md and Notes.md
Jonatan Werpers <jonatan@werpers.com>
parents: 427
diff changeset
155 This would mean no bounds checking in applys, however any indexing that they do would be boundschecked. The only loss would be readability of errors. But users aren't really supposed to call apply directly anyway.
8f9b3eac128a Add a bunch of stuff to TODO.md and Notes.md
Jonatan Werpers <jonatan@werpers.com>
parents: 427
diff changeset
156
8f9b3eac128a Add a bunch of stuff to TODO.md and Notes.md
Jonatan Werpers <jonatan@werpers.com>
parents: 427
diff changeset
157 Preferably dimensions and sizes should be checked when lazy objects are created, for example TensorApplication, TensorComposition and so on. If dimension checks decreases performance we can make them skippable later.
8f9b3eac128a Add a bunch of stuff to TODO.md and Notes.md
Jonatan Werpers <jonatan@werpers.com>
parents: 427
diff changeset
158
410
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
159 ## Vector valued grid functions
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
160 Från slack konversation:
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
161
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
162 Jonatan Werpers:
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
163 Med vektorvärda gridfunktioner vill vi ju fortfarande att grid funktionen ska vara till exempel AbstractArray{LitenVektor,2}
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
164 Och att man ska kunna göra allt man vill med LitenVektor
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
165 typ addera, jämföra osv
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
166 Och då borde points returnera AbstractArray{LitenVektor{Float,2},2} för ett 2d nät
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
167 Men det kanske bara ska vara Static arrays?
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
168
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
169 Vidar Stiernström:
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
170 Ja, jag vet inte riktigt vad som är en rimlig representation
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
171 Du menar en vektor av static arrays då?
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
172
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
173 Jonatan Werpers:
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
174 Ja, att LitenVektor är en StaticArray
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
175
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
176 Vidar Stiernström:
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
177 Tuplar känns typ rätt inuitivt för att representera värdet i en punkt
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
178 men
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
179 det suger att man inte har + och - för dem
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
180
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
181 Jonatan Werpers:
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
182 Ja precis
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
183
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
184 Vidar Stiernström:
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
185 så kanske är bra med static arrays i detta fall
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
186
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
187 Jonatan Werpers:
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
188 Man vill ju kunna köra en Operator rakt på och vara klar eller?
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
189
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
190 Vidar Stiernström:
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
191 Har inte alls tänkt på hur det vi gör funkar mot vektorvärda funktioner
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
192 men känns som staticarrays är hur man vill göra det
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
193 tuplar är ju immutable också
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
194 blir jobbigt om man bara agerar på en komponent då
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
195
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
196 Jonatan Werpers:
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
197 Hm…
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
198 Tål att tänkas på
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
199 Men det lär ju bli mer indirektion med mutables eller?
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
200 Hur fungerar det?
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
201 Det finns ju hur som helst både SVector och MVector i StaticArrays
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
202
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
203 Vidar Stiernström:
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
204 När vi jobbat i c/c++ och kollat runt lite hur man brukar göra så lagrar man i princip alla sina obekanta i en lång vektor och så får man specificera i funktioerna vilken komponent man agerar på och till vilken man skriver
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
205 så man lagrar grejer enl: w = [u1, v1, u2, v2, …] i 1D.
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
206 Men alltså har ingen aning hur julia hanterar detta
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
207
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
208 Jonatan Werpers:
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
209 Det vi är ute efter kanske är att en grid funcktion är en AbstractArray{T,2} where T<:NågotSomViKanRäknaMed
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
210 Och så får den typen var lite vad som helst.
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
211
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
212 Vidar Stiernström:
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
213 Tror det kan vara farligt att ha nåt som är AbstractArray{LitenArray{NDof},Dim}
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
214 Jag gissar att det kompilatorn vill ha är en stor array med doubles
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
215
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
216 Jonatan Werpers:
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
217 Och sen är det upp till den som använder grejerna att vara smart
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
218 Vill man vara trixig kan man väl då imlementera SuperHaxxorGridFunction <: AbstractArray{Array{…},2} som lagrar allt linjärt eller något sånt
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
219 Det kommer väl lösa sig när man börjar implementera vektorvärda saker
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
220 Euler nästa!
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
221 New
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
222 Vidar Stiernström:
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
223 Det vore skönt att inte behöva skriva såhär varje gång man testar mot en tupel :smile: @test [gp[i]...] ≈ [p[i]...] atol=5e-13
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
224
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
225 Jonatan Werpers:
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
226 https://github.com/JuliaArrays/ArraysOfArrays.jl
26e186b565b3 Add some notes about vector valued grid functions
Jonatan Werpers <jonatan@werpers.com>
parents: 379
diff changeset
227 https://github.com/jw3126/Setfield.jl
423
10a67ac48d6e Add some notes
Jonatan Werpers <jonatan@werpers.com>
parents: 410
diff changeset
228
10a67ac48d6e Add some notes
Jonatan Werpers <jonatan@werpers.com>
parents: 410
diff changeset
229 ### Test-applikationer
10a67ac48d6e Add some notes
Jonatan Werpers <jonatan@werpers.com>
parents: 410
diff changeset
230 div och grad operationer
10a67ac48d6e Add some notes
Jonatan Werpers <jonatan@werpers.com>
parents: 410
diff changeset
231
606
8f9b3eac128a Add a bunch of stuff to TODO.md and Notes.md
Jonatan Werpers <jonatan@werpers.com>
parents: 427
diff changeset
232 Enligt Wikipedia verkar det som att `∇⋅` agerar på första dimensionen av ett tensor fält och `div()` på sista.
8f9b3eac128a Add a bunch of stuff to TODO.md and Notes.md
Jonatan Werpers <jonatan@werpers.com>
parents: 427
diff changeset
233 Om man generaliserar kanske `∇` i så fall bara lägger till en dimension i början.
8f9b3eac128a Add a bunch of stuff to TODO.md and Notes.md
Jonatan Werpers <jonatan@werpers.com>
parents: 427
diff changeset
234
8f9b3eac128a Add a bunch of stuff to TODO.md and Notes.md
Jonatan Werpers <jonatan@werpers.com>
parents: 427
diff changeset
235 Kan vi implementera `⋅`(\cdot) så att de fungerar som man vill för både tensor-fält och tensor-operatorer?
427
1c41f4fd3e61 Add some notes and todos
Jonatan Werpers <jonatan@werpers.com>
parents: 423
diff changeset
236
606
8f9b3eac128a Add a bunch of stuff to TODO.md and Notes.md
Jonatan Werpers <jonatan@werpers.com>
parents: 427
diff changeset
237 Är `∇` ett tensor-fält av tensor-operatorer? Vad är ett tensor-fält i vår kod? Är det en special-fall av en tensor-mapping?
427
1c41f4fd3e61 Add some notes and todos
Jonatan Werpers <jonatan@werpers.com>
parents: 423
diff changeset
238
606
8f9b3eac128a Add a bunch of stuff to TODO.md and Notes.md
Jonatan Werpers <jonatan@werpers.com>
parents: 427
diff changeset
239 ### Grid-funktionen
8f9b3eac128a Add a bunch of stuff to TODO.md and Notes.md
Jonatan Werpers <jonatan@werpers.com>
parents: 427
diff changeset
240 Grid-funktionon har typen `AbstractArray{T,2} where T`.
8f9b3eac128a Add a bunch of stuff to TODO.md and Notes.md
Jonatan Werpers <jonatan@werpers.com>
parents: 427
diff changeset
241 `T` kan vara lite vad som helst, tillexemel en SVector eller Array, eller tuple. TensorOperatorerna bryr sig inte om exakt vad det är, mer än att typen måste stödja de operationer som operatorn använder.
8f9b3eac128a Add a bunch of stuff to TODO.md and Notes.md
Jonatan Werpers <jonatan@werpers.com>
parents: 427
diff changeset
242
8f9b3eac128a Add a bunch of stuff to TODO.md and Notes.md
Jonatan Werpers <jonatan@werpers.com>
parents: 427
diff changeset
243 En nackdel kan vara hur man ska få ut gridfunktionen för tex andra komponenten.
427
1c41f4fd3e61 Add some notes and todos
Jonatan Werpers <jonatan@werpers.com>
parents: 423
diff changeset
244
1c41f4fd3e61 Add some notes and todos
Jonatan Werpers <jonatan@werpers.com>
parents: 423
diff changeset
245 Syntax:
1c41f4fd3e61 Add some notes and todos
Jonatan Werpers <jonatan@werpers.com>
parents: 423
diff changeset
246 ```
606
8f9b3eac128a Add a bunch of stuff to TODO.md and Notes.md
Jonatan Werpers <jonatan@werpers.com>
parents: 427
diff changeset
247 f(x̄) = x̄
8f9b3eac128a Add a bunch of stuff to TODO.md and Notes.md
Jonatan Werpers <jonatan@werpers.com>
parents: 427
diff changeset
248 gf = evalOn(g, f)
8f9b3eac128a Add a bunch of stuff to TODO.md and Notes.md
Jonatan Werpers <jonatan@werpers.com>
parents: 427
diff changeset
249 gf[2,3] # x̄ för en viss gridpunkt
8f9b3eac128a Add a bunch of stuff to TODO.md and Notes.md
Jonatan Werpers <jonatan@werpers.com>
parents: 427
diff changeset
250 gf[2,3][2] # x̄[2] för en viss gridpunkt
427
1c41f4fd3e61 Add some notes and todos
Jonatan Werpers <jonatan@werpers.com>
parents: 423
diff changeset
251 ```
1c41f4fd3e61 Add some notes and todos
Jonatan Werpers <jonatan@werpers.com>
parents: 423
diff changeset
252
606
8f9b3eac128a Add a bunch of stuff to TODO.md and Notes.md
Jonatan Werpers <jonatan@werpers.com>
parents: 427
diff changeset
253 Note: Behöver bestämma om eval on skickar in `x̄` eller `x̄...` till `f`. Eller om man kan stödja båda.
427
1c41f4fd3e61 Add some notes and todos
Jonatan Werpers <jonatan@werpers.com>
parents: 423
diff changeset
254
606
8f9b3eac128a Add a bunch of stuff to TODO.md and Notes.md
Jonatan Werpers <jonatan@werpers.com>
parents: 427
diff changeset
255 ### Tensor operatorer
8f9b3eac128a Add a bunch of stuff to TODO.md and Notes.md
Jonatan Werpers <jonatan@werpers.com>
parents: 427
diff changeset
256 Vi kan ha tensor-operatorer som agerar på ett skalärt fält och ger ett vektorfält eller tensorfält.
8f9b3eac128a Add a bunch of stuff to TODO.md and Notes.md
Jonatan Werpers <jonatan@werpers.com>
parents: 427
diff changeset
257 Vi kan också ha tensor-operatorer som agerar på ett vektorfält eller tensorfält och ger ett skalärt fält.
8f9b3eac128a Add a bunch of stuff to TODO.md and Notes.md
Jonatan Werpers <jonatan@werpers.com>
parents: 427
diff changeset
258
8f9b3eac128a Add a bunch of stuff to TODO.md and Notes.md
Jonatan Werpers <jonatan@werpers.com>
parents: 427
diff changeset
259 TBD: Just nu gör `apply_transpose` antagandet att domän-typen är samma som range-typen. Det behöver vi på något sätt bryta. Ett alternativ är låta en TensorMapping ha `T_domain` och `T_range` istället för bara `T`. Känns dock lite grötigt. Ett annat alternativ skulle vara någon typ av trait för transpose? Den skulle kunna innehålla typen som transponatet agerar på? Vet inte om det fungerar dock.
8f9b3eac128a Add a bunch of stuff to TODO.md and Notes.md
Jonatan Werpers <jonatan@werpers.com>
parents: 427
diff changeset
260
8f9b3eac128a Add a bunch of stuff to TODO.md and Notes.md
Jonatan Werpers <jonatan@werpers.com>
parents: 427
diff changeset
261 TBD: Vad är målet med `T`-parametern för en TensorMapping? Om vi vill kunna applicera en difference operator på vad som helst kan man inte anta att en `TensorMapping{T}` bara agerar på instanser av `T`.
427
1c41f4fd3e61 Add some notes and todos
Jonatan Werpers <jonatan@werpers.com>
parents: 423
diff changeset
262
606
8f9b3eac128a Add a bunch of stuff to TODO.md and Notes.md
Jonatan Werpers <jonatan@werpers.com>
parents: 427
diff changeset
263 Man kan implementera `∇` som en tensormapping som agerar på T och returnerar `StaticVector{N,T} where N`.
8f9b3eac128a Add a bunch of stuff to TODO.md and Notes.md
Jonatan Werpers <jonatan@werpers.com>
parents: 427
diff changeset
264 (Man skulle eventuellt också kunna låta den agera på `StaticMatrix{N,T,D} where N` och returnera `StaticMatrix{M,T,D+1}`. Frågan är om man vinner något på det...)
8f9b3eac128a Add a bunch of stuff to TODO.md and Notes.md
Jonatan Werpers <jonatan@werpers.com>
parents: 427
diff changeset
265
8f9b3eac128a Add a bunch of stuff to TODO.md and Notes.md
Jonatan Werpers <jonatan@werpers.com>
parents: 427
diff changeset
266 Skulle kunna ha en funktion `range_type(::TensorMapping, ::Type{domain_type})`
8f9b3eac128a Add a bunch of stuff to TODO.md and Notes.md
Jonatan Werpers <jonatan@werpers.com>
parents: 427
diff changeset
267
8f9b3eac128a Add a bunch of stuff to TODO.md and Notes.md
Jonatan Werpers <jonatan@werpers.com>
parents: 427
diff changeset
268 Kanske kan man implementera `⋅(tm::TensorMapping{R,D}, v::AbstractArray{T,D})` där T är en AbstractArray, tm på något sätt har komponenter, lika många som T har element.
8f9b3eac128a Add a bunch of stuff to TODO.md and Notes.md
Jonatan Werpers <jonatan@werpers.com>
parents: 427
diff changeset
269
8f9b3eac128a Add a bunch of stuff to TODO.md and Notes.md
Jonatan Werpers <jonatan@werpers.com>
parents: 427
diff changeset
270 ### Ratade alternativ:
8f9b3eac128a Add a bunch of stuff to TODO.md and Notes.md
Jonatan Werpers <jonatan@werpers.com>
parents: 427
diff changeset
271
8f9b3eac128a Add a bunch of stuff to TODO.md and Notes.md
Jonatan Werpers <jonatan@werpers.com>
parents: 427
diff changeset
272
8f9b3eac128a Add a bunch of stuff to TODO.md and Notes.md
Jonatan Werpers <jonatan@werpers.com>
parents: 427
diff changeset
273 #### 2.AbstractArray{T,2+1} where T (NOPE!)
8f9b3eac128a Add a bunch of stuff to TODO.md and Notes.md
Jonatan Werpers <jonatan@werpers.com>
parents: 427
diff changeset
274 Blir inte den här. Bryter mot alla tankar om hur grid funktioner ska fungera. Om de tillåts ha en annan dimension än nätet blir allt hemskt.
8f9b3eac128a Add a bunch of stuff to TODO.md and Notes.md
Jonatan Werpers <jonatan@werpers.com>
parents: 427
diff changeset
275
427
1c41f4fd3e61 Add some notes and todos
Jonatan Werpers <jonatan@werpers.com>
parents: 423
diff changeset
276 Man låter helt enkelt arrayen ha en extra dimension. En fördel är att man har en väldigt "native" typ. En nackdel kan vara att det eventuellt blir rörigt vilken dimension olika operatorer ska agera på. I värsta fall behöver vi "kroneckra in" de tillagda dimensionerna. Vektorfältets index kommer också att bli det första eftersom vi vill att de ska lagras kontinuerligt i minnet pga chachen. (Går kanske att lösa med en custom typ men då krånglar man till det för sig). En fördel skulle vara att man enkelt får ut olika komponenter.
1c41f4fd3e61 Add some notes and todos
Jonatan Werpers <jonatan@werpers.com>
parents: 423
diff changeset
277
1c41f4fd3e61 Add some notes and todos
Jonatan Werpers <jonatan@werpers.com>
parents: 423
diff changeset
278 Syntax:
1c41f4fd3e61 Add some notes and todos
Jonatan Werpers <jonatan@werpers.com>
parents: 423
diff changeset
279 ```
1c41f4fd3e61 Add some notes and todos
Jonatan Werpers <jonatan@werpers.com>
parents: 423
diff changeset
280 gf = eval_on_grid(g,f)
1c41f4fd3e61 Add some notes and todos
Jonatan Werpers <jonatan@werpers.com>
parents: 423
diff changeset
281 gf[:,2,3] # Hela vektorn för en gridpunkt
1c41f4fd3e61 Add some notes and todos
Jonatan Werpers <jonatan@werpers.com>
parents: 423
diff changeset
282 gf[2,2,3] # Andra komponenten av vektor fältet i en punkt.
1c41f4fd3e61 Add some notes and todos
Jonatan Werpers <jonatan@werpers.com>
parents: 423
diff changeset
283 gf[2,:,:] #
1c41f4fd3e61 Add some notes and todos
Jonatan Werpers <jonatan@werpers.com>
parents: 423
diff changeset
284 ```
1c41f4fd3e61 Add some notes and todos
Jonatan Werpers <jonatan@werpers.com>
parents: 423
diff changeset
285
1c41f4fd3e61 Add some notes and todos
Jonatan Werpers <jonatan@werpers.com>
parents: 423
diff changeset
286 ### Evaluering av funktioner på nät
1c41f4fd3e61 Add some notes and todos
Jonatan Werpers <jonatan@werpers.com>
parents: 423
diff changeset
287 Hur ska man skriva funktioner som evalueras på nätet? `f(x,y) = ...` eller `f(x̄) = ...`? Eller båda? Kan eval_on_grid se skillnad eller får användaren specificera?
1c41f4fd3e61 Add some notes and todos
Jonatan Werpers <jonatan@werpers.com>
parents: 423
diff changeset
288
1c41f4fd3e61 Add some notes and todos
Jonatan Werpers <jonatan@werpers.com>
parents: 423
diff changeset
289 ```
1c41f4fd3e61 Add some notes and todos
Jonatan Werpers <jonatan@werpers.com>
parents: 423
diff changeset
290 f(x,y) = [x^2, y^2]
1c41f4fd3e61 Add some notes and todos
Jonatan Werpers <jonatan@werpers.com>
parents: 423
diff changeset
291 f(x̄) = [x̄[1]^2, x̄[2]^2]
1c41f4fd3e61 Add some notes and todos
Jonatan Werpers <jonatan@werpers.com>
parents: 423
diff changeset
292 ```
1c41f4fd3e61 Add some notes and todos
Jonatan Werpers <jonatan@werpers.com>
parents: 423
diff changeset
293
1c41f4fd3e61 Add some notes and todos
Jonatan Werpers <jonatan@werpers.com>
parents: 423
diff changeset
294 Påverkas detta av hur vi förväntar oss kunna skapa lata gridfunktioner?
1c41f4fd3e61 Add some notes and todos
Jonatan Werpers <jonatan@werpers.com>
parents: 423
diff changeset
295
1c41f4fd3e61 Add some notes and todos
Jonatan Werpers <jonatan@werpers.com>
parents: 423
diff changeset
296 ### Komponenter som gridfunktioner
606
8f9b3eac128a Add a bunch of stuff to TODO.md and Notes.md
Jonatan Werpers <jonatan@werpers.com>
parents: 427
diff changeset
297 En viktig operation för vektor fält är att kunna få ut komponenter som grid-funktioner. Detta behöver antagligen kunna ske lazy.
427
1c41f4fd3e61 Add some notes and todos
Jonatan Werpers <jonatan@werpers.com>
parents: 423
diff changeset
298 Det finns ett par olika lösningar:
1c41f4fd3e61 Add some notes and todos
Jonatan Werpers <jonatan@werpers.com>
parents: 423
diff changeset
299 * Implementera en egen typ av view som tar hand om detta. Eller Accessors.jl?
1c41f4fd3e61 Add some notes and todos
Jonatan Werpers <jonatan@werpers.com>
parents: 423
diff changeset
300 * Använda en TensorMapping
1c41f4fd3e61 Add some notes and todos
Jonatan Werpers <jonatan@werpers.com>
parents: 423
diff changeset
301 * Någon typ av lazy-broadcast
1c41f4fd3e61 Add some notes and todos
Jonatan Werpers <jonatan@werpers.com>
parents: 423
diff changeset
302 * En lazy array som applicerar en funktion för varje element.
1c41f4fd3e61 Add some notes and todos
Jonatan Werpers <jonatan@werpers.com>
parents: 423
diff changeset
303
1c41f4fd3e61 Add some notes and todos
Jonatan Werpers <jonatan@werpers.com>
parents: 423
diff changeset
304 Skulle vara en fördel om det är hyffsat generiskt så att en eventuell användare kan utöka det enkelt om de har någon egen exotisk typ. Eller ska man vila helt på
1c41f4fd3e61 Add some notes and todos
Jonatan Werpers <jonatan@werpers.com>
parents: 423
diff changeset
305
1c41f4fd3e61 Add some notes and todos
Jonatan Werpers <jonatan@werpers.com>
parents: 423
diff changeset
306 Syntax:
1c41f4fd3e61 Add some notes and todos
Jonatan Werpers <jonatan@werpers.com>
parents: 423
diff changeset
307 ```
1c41f4fd3e61 Add some notes and todos
Jonatan Werpers <jonatan@werpers.com>
parents: 423
diff changeset
308 gf = eval(...)
1c41f4fd3e61 Add some notes and todos
Jonatan Werpers <jonatan@werpers.com>
parents: 423
diff changeset
309 component(gf,2) # Andra komponenten av en vektor
1c41f4fd3e61 Add some notes and todos
Jonatan Werpers <jonatan@werpers.com>
parents: 423
diff changeset
310 component(gf,2,3) # (2,3) elementet av en matris
1c41f4fd3e61 Add some notes and todos
Jonatan Werpers <jonatan@werpers.com>
parents: 423
diff changeset
311 component(gf,:,2) # Andra kolumnen av en matris
1c41f4fd3e61 Add some notes and todos
Jonatan Werpers <jonatan@werpers.com>
parents: 423
diff changeset
312 @ourview gf[:,:][2]
1c41f4fd3e61 Add some notes and todos
Jonatan Werpers <jonatan@werpers.com>
parents: 423
diff changeset
313 ```
683
841ca12f3359 Add notes about embedded grids
Jonatan Werpers <jonatan@werpers.com>
parents: 654
diff changeset
314
841ca12f3359 Add notes about embedded grids
Jonatan Werpers <jonatan@werpers.com>
parents: 654
diff changeset
315 ## Grids embedded in higher dimensions
841ca12f3359 Add notes about embedded grids
Jonatan Werpers <jonatan@werpers.com>
parents: 654
diff changeset
316
841ca12f3359 Add notes about embedded grids
Jonatan Werpers <jonatan@werpers.com>
parents: 654
diff changeset
317 For grids generated by asking for boundary grids for a regular grid, it would
841ca12f3359 Add notes about embedded grids
Jonatan Werpers <jonatan@werpers.com>
parents: 654
diff changeset
318 make sense if these grids knew they were embedded in a higher dimension. They
841ca12f3359 Add notes about embedded grids
Jonatan Werpers <jonatan@werpers.com>
parents: 654
diff changeset
319 would return coordinates in the full room. This would make sense when
841ca12f3359 Add notes about embedded grids
Jonatan Werpers <jonatan@werpers.com>
parents: 654
diff changeset
320 drawing points for example, or when evaluating functions on the boundary.
841ca12f3359 Add notes about embedded grids
Jonatan Werpers <jonatan@werpers.com>
parents: 654
diff changeset
321
841ca12f3359 Add notes about embedded grids
Jonatan Werpers <jonatan@werpers.com>
parents: 654
diff changeset
322 Implementation of this is an issue that requires some thought. Adding an extra
841ca12f3359 Add notes about embedded grids
Jonatan Werpers <jonatan@werpers.com>
parents: 654
diff changeset
323 "Embedded" type for each grid would make it easy to understand each type but
841ca12f3359 Add notes about embedded grids
Jonatan Werpers <jonatan@werpers.com>
parents: 654
diff changeset
324 contribute to "type bloat". On the other hand adapting existing types to
841ca12f3359 Add notes about embedded grids
Jonatan Werpers <jonatan@werpers.com>
parents: 654
diff changeset
325 handle embeddedness would complicate the now very simple grid types. Are there
841ca12f3359 Add notes about embedded grids
Jonatan Werpers <jonatan@werpers.com>
parents: 654
diff changeset
326 other ways of doing the implentation?
718
05d8ea88c690 Add note about performance testin
Jonatan Werpers <jonatan@werpers.com>
parents: 683
diff changeset
327
05d8ea88c690 Add note about performance testin
Jonatan Werpers <jonatan@werpers.com>
parents: 683
diff changeset
328 ## Performance measuring
05d8ea88c690 Add note about performance testin
Jonatan Werpers <jonatan@werpers.com>
parents: 683
diff changeset
329 We should be measuring performance early. How does our effective cpu and memory bandwidth utilization compare to peak performance?
05d8ea88c690 Add note about performance testin
Jonatan Werpers <jonatan@werpers.com>
parents: 683
diff changeset
330
05d8ea88c690 Add note about performance testin
Jonatan Werpers <jonatan@werpers.com>
parents: 683
diff changeset
331 We should make these test simple to run for any solver.
05d8ea88c690 Add note about performance testin
Jonatan Werpers <jonatan@werpers.com>
parents: 683
diff changeset
332
05d8ea88c690 Add note about performance testin
Jonatan Werpers <jonatan@werpers.com>
parents: 683
diff changeset
333 See [this talk](https://www.youtube.com/watch?v=vPsfZUqI4_0) for some simple ideas for defining effecive memory usage and some comparison with peak performance.
05d8ea88c690 Add note about performance testin
Jonatan Werpers <jonatan@werpers.com>
parents: 683
diff changeset
334
05d8ea88c690 Add note about performance testin
Jonatan Werpers <jonatan@werpers.com>
parents: 683
diff changeset
335
761
d7d030f8f708 Add note about adjoint as trait
Jonatan Werpers <jonatan@werpers.com>
parents: 718
diff changeset
336 ## Adjoint as a trait on the sbp_operator level?
d7d030f8f708 Add note about adjoint as trait
Jonatan Werpers <jonatan@werpers.com>
parents: 718
diff changeset
337
d7d030f8f708 Add note about adjoint as trait
Jonatan Werpers <jonatan@werpers.com>
parents: 718
diff changeset
338 It would be nice to have a way of refering to adjoints with resepct to the sbp-inner-product.
d7d030f8f708 Add note about adjoint as trait
Jonatan Werpers <jonatan@werpers.com>
parents: 718
diff changeset
339 If it was possible you could reduce the number of times you have to deal with the inner product matrix.
d7d030f8f708 Add note about adjoint as trait
Jonatan Werpers <jonatan@werpers.com>
parents: 718
diff changeset
340
d7d030f8f708 Add note about adjoint as trait
Jonatan Werpers <jonatan@werpers.com>
parents: 718
diff changeset
341 Since the LazyOperators package is sort of implementing matrix-free matrices there is no concept of inner products there at the moment. It seems to complicate large parts of the package if this was included there.
d7d030f8f708 Add note about adjoint as trait
Jonatan Werpers <jonatan@werpers.com>
parents: 718
diff changeset
342
d7d030f8f708 Add note about adjoint as trait
Jonatan Werpers <jonatan@werpers.com>
parents: 718
diff changeset
343 A different approach would be to include it as a trait for operators so that you can specify what the adjoint for that operator is.
d7d030f8f708 Add note about adjoint as trait
Jonatan Werpers <jonatan@werpers.com>
parents: 718
diff changeset
344
d7d030f8f708 Add note about adjoint as trait
Jonatan Werpers <jonatan@werpers.com>
parents: 718
diff changeset
345
775
1d166e7a9deb Add a note about the name of VolumeOperator
Jonatan Werpers <jonatan@werpers.com>
parents: 761
diff changeset
346 ## Name of the `VolumeOperator` type for constant stencils
1d166e7a9deb Add a note about the name of VolumeOperator
Jonatan Werpers <jonatan@werpers.com>
parents: 761
diff changeset
347 It seems that the name is too general. The name of the method `volume_operator` makes sense. It should return different types of `TensorMapping` specialized for the grid. A suggetion for a better name is `ConstantStencilVolumeOperator`