Incidence graph: Difference between revisions
No edit summary |
No edit summary |
||
| (9 intermediate revisions by the same user not shown) | |||
| Line 1: | Line 1: | ||
<div class="pageexperimental">''This page documents an '''experimental''' feature. You can help out by '''testing it''' and <span class="plainlinks">[http:// | <div class="pageexperimental">''This page documents an '''experimental''' feature. You can help out by '''testing it''' and <span class="plainlinks">[http://bugs.ascend4.org/ reporting]</span> your experiences.''</div> | ||
You can create an incidence graph and view it on-screen if your copy of ASCEND is linked to [http://www.graphviz.org GraphViz]. We have currently only tested this on Linux, but it should transfer to Windows without too many problems. | You can create an incidence graph and view it on-screen if your copy of ASCEND is linked to [http://www.graphviz.org GraphViz]. We have currently only tested this on Linux, but it should transfer to Windows without too many problems. | ||
| Line 5: | Line 5: | ||
To show the incidence graph, use the ''Tools->Incidence Graph'' menu option in the PyGTK GUI. | To show the incidence graph, use the ''Tools->Incidence Graph'' menu option in the PyGTK GUI. | ||
For the simple {{src|models/johnpye/ | For the simple {{src|models/johnpye/reynolds.a4c}} model, the following incidence graph results: | ||
[[Image:Incidence-graph-reynolds.png| | [[Image:Incidence-graph-reynolds.png|200px|thumb|none|Incidence graph for {{src|models/johnpye/reynolds.a4c}}.]] | ||
A more complex example is this one, which | A more complex example is this one, which shows the calculation of the properties of CO2 when pressure and vapour fraction are specified, from equations in terms of pressure and enthalpy ({{src|models/johnpye/fprops/rankine_fprops.a4c}}, see <tt>stream_state_test_co2</tt>): | ||
[[Image:Rankine-incidence-graph.png| | [[Image:Rankine-incidence-graph.png|300px|thumb|none|Incidence graph for iterative calculation of CO2 properties ({{src|models/johnpye/fprops/rankine_fprops.a4c}}). Note the upward arrows, which imply that a simply single-pass calculation is not possible for this problem.]] | ||
In both of these diagrams, the fixed variables are shown in green; all other variables are shown in boxes, and equations are shown in ellipses. | In both of these diagrams, the fixed variables are shown in green; all other variables are shown in boxes, and equations are shown in ellipses. | ||
There are limits to how complex a model is allowed to be when creating these incidence graphs. ASCEND will build these images up to about 400 variables, but such diagrams are really so complex as to be not very useful. | |||
Various improvements to the graph might be possible by ranking the variables, so that they appear down the page in the order that ASCEND solves them. | |||
== See also == | == See also == | ||
Latest revision as of 01:33, 26 November 2017
You can create an incidence graph and view it on-screen if your copy of ASCEND is linked to GraphViz. We have currently only tested this on Linux, but it should transfer to Windows without too many problems.
To show the incidence graph, use the Tools->Incidence Graph menu option in the PyGTK GUI.
For the simple models/johnpye/reynolds.a4c model, the following incidence graph results:

A more complex example is this one, which shows the calculation of the properties of CO2 when pressure and vapour fraction are specified, from equations in terms of pressure and enthalpy (models/johnpye/fprops/rankine_fprops.a4c, see stream_state_test_co2):

In both of these diagrams, the fixed variables are shown in green; all other variables are shown in boxes, and equations are shown in ellipses.
There are limits to how complex a model is allowed to be when creating these incidence graphs. ASCEND will build these images up to about 400 variables, but such diagrams are really so complex as to be not very useful.
Various improvements to the graph might be possible by ranking the variables, so that they appear down the page in the order that ASCEND solves them.