1 | ANUGA Future Directions (2005-2006) |
---|
2 | |
---|
3 | |
---|
4 | -------------------------------------------------- |
---|
5 | Draft of known issues regarding the ANUGA software |
---|
6 | -------------------------------------------------- |
---|
7 | |
---|
8 | |
---|
9 | ---------------- |
---|
10 | Numerical issues |
---|
11 | |
---|
12 | CFL condition (centroid-midpt distances versus inscribed circle ? or error estimator) |
---|
13 | Gradient limiters |
---|
14 | Forcing terms: |
---|
15 | Friction of Mud |
---|
16 | Viscosity |
---|
17 | |
---|
18 | |
---|
19 | |
---|
20 | -------------------- |
---|
21 | Functionality issues |
---|
22 | |
---|
23 | Mesh generation: |
---|
24 | Through graphical generator |
---|
25 | Through scripts |
---|
26 | Alternative triangulators (Strang, ?.) |
---|
27 | alpha-shapes |
---|
28 | |
---|
29 | refactoring triangle interface |
---|
30 | outputting a ?triangle? readable file (from pmesh?) |
---|
31 | |
---|
32 | from pmesh issues file: |
---|
33 | Main issues (TBA) |
---|
34 | |
---|
35 | Georeferencing |
---|
36 | Code works in UTM coordinates. Zone boundaries cannot be crossed |
---|
37 | Could code be cast in terms of lat/lons? |
---|
38 | |
---|
39 | least_squares, use of quad-tree may not use all datapoints |
---|
40 | (currently fixed by expand search, need algorithm for searching neighbouring treenodes) |
---|
41 | |
---|
42 | Give feedback when |
---|
43 | points outside mesh |
---|
44 | mesh outside points boundary |
---|
45 | |
---|
46 | use sparse matrix (CSR) format in ?build_interpolation_matrix_A? |
---|
47 | simplify geo-referencing |
---|
48 | |
---|
49 | Checkpointing (largely done but needs to be revisited) |
---|
50 | |
---|
51 | Ability to create additional quantities and initialise arbitrarily |
---|
52 | |
---|
53 | FIXMEs (TBA) |
---|
54 | |
---|
55 | |
---|
56 | File management (msh, pts, dem, ..) ? introduce the notion of a project |
---|
57 | - reduce # of methods that read/write sww files (Yes) |
---|
58 | |
---|
59 | |
---|
60 | ----------------- |
---|
61 | Efficiency issues |
---|
62 | |
---|
63 | Parallelisation |
---|
64 | |
---|
65 | Edgefluxes are currently computed twice (f01 = -f10) |
---|
66 | (This has been fixed by Matt Hardy) |
---|
67 | |
---|
68 | Increase Timesteps (Implicit techniques?) |
---|
69 | |
---|
70 | Optimise creation and importing of domains (C-code and caching) |
---|
71 | |
---|
72 | |
---|
73 | ------------ |
---|
74 | Architecture |
---|
75 | |
---|
76 | Settle dataflows and formats (pts, dem, msh, sww, etc) |
---|
77 | |
---|
78 | ?finaltime? to be renamed ?duration? |
---|
79 | |
---|
80 | |
---|
81 | ---- |
---|
82 | API: |
---|
83 | |
---|
84 | Suggestion for set_quantity |
---|
85 | set_quantity(name, X, location, region) |
---|
86 | where |
---|
87 | name: Name of quantity |
---|
88 | X: |
---|
89 | -Compatible list, |
---|
90 | -Numeric array, |
---|
91 | -constant |
---|
92 | -callable object (f: x, y -> z) where x,y,z are arrays |
---|
93 | - inline |
---|
94 | - file functions |
---|
95 | - polygon functions |
---|
96 | -another quantity Q or an expression of the form |
---|
97 | a*Q+b*R+c, where |
---|
98 | a,b,c are scalars, arrays or quantities |
---|
99 | Q,R are quantities (or 1?) |
---|
100 | - list of points with attribute values (use LS) |
---|
101 | - pts filename (use LS and caching) |
---|
102 | - general expression to be parsed |
---|
103 | |
---|
104 | location: Where values are to be stored. |
---|
105 | Permissible options are: vertices, edges, centroid, unique_vertices |
---|
106 | region: Identify subset of triangles. Permissible values are |
---|
107 | - tag name (refers to tagged region) |
---|
108 | - indices (refers to specific triangles) |
---|
109 | - polygon (identifies region) |
---|
110 | (incorporate uniqueness/non-uniqueness) |
---|
111 | |
---|
112 | |
---|
113 | Making methods private, using _private |
---|
114 | Or write the API in a separate module. |
---|
115 | |
---|
116 | |
---|
117 | Make stage appear as any other quantity: |
---|
118 | Either |
---|
119 | 1: Make stage a subclass of quantity having knowledge of elevation and a special |
---|
120 | limiter (or more limiters) |
---|
121 | Or |
---|
122 | 2: Equip each quantity with a limiter class |
---|
123 | (Con: A limiter for stage should never be applied to any other quantity) |
---|
124 | |
---|
125 | |
---|
126 | Also, investigate if Quantity and Conserved_quantity should be one class (Steve). |
---|
127 | |
---|
128 | Finally, reconcile the optimised gradient limiter of Matt's with the more general framework |
---|
129 | |
---|
130 | |
---|
131 | |
---|
132 | ------ |
---|
133 | Other: |
---|
134 | |
---|
135 | Should we remove Python code superseded by C-extensions? |
---|
136 | Pros: Leaner code and no risk |
---|
137 | Cons: Less readable algorithms |
---|
138 | |
---|
139 | Move py code into files such as quantity_ext.py and have python wrappers with doc strings for all functions. Name the code in extensions _ext and conditional import as usual. |
---|
140 | |
---|
141 | |
---|
142 | ------------- |
---|
143 | Visualisation |
---|
144 | |
---|
145 | Swollen: |
---|
146 | Export to movie |
---|
147 | Use maps |
---|
148 | Colourcoding |
---|
149 | z-scale |
---|
150 | |
---|
151 | Matplotlib: |
---|
152 | Realtime and ?post mortem? generation of colour coded maps with contourlines. |
---|
153 | |
---|
154 | Visual Python: |
---|
155 | Steves new and improved tool |
---|
156 | |
---|
157 | |
---|
158 | |
---|
159 | ----------- |
---|
160 | Development |
---|
161 | |
---|
162 | Restructuring and moving svn repository |
---|
163 | |
---|
164 | Flatten directory structure |
---|
165 | |
---|
166 | Apply buildbot |
---|
167 | |
---|
168 | Use real bugtracking/project management tool |
---|
169 | (see http://www.generalconcepts.com/resources/tracking, plone, basecamp,? |
---|
170 | Try out TRAC which integrates into subversion |
---|
171 | |
---|
172 | Zeus? |
---|
173 | |
---|
174 | |
---|
175 | -------------- |
---|
176 | Documentation |
---|
177 | |
---|
178 | Hire technical writer to produce |
---|
179 | - Getting started |
---|
180 | - User guide |
---|
181 | - Reference manual ? semi automated |
---|
182 | |
---|
183 | Mathematical model description (Steve & Chris) |
---|
184 | |
---|
185 | |
---|
186 | ----------------- |
---|
187 | Validation and QA |
---|
188 | |
---|
189 | Merimbula (Steve & Chris) |
---|
190 | PNG Landslide study or watertank data (Adrian?) |
---|
191 | BOM |
---|
192 | SMEC |
---|
193 | CSIRO |
---|
194 | |
---|
195 | |
---|
196 | -------- |
---|
197 | Naming |
---|
198 | |
---|
199 | The collection as well as individual modules may benefit from better names. |
---|
200 | |
---|
201 | |
---|
202 | -------- |
---|
203 | Release |
---|
204 | |
---|
205 | Search for appropriate procedure for OSS release (Ole is onto that with AGIMO) |
---|
206 | |
---|
207 | |
---|
208 | |
---|