1 | <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN"> |
---|
2 | <html> |
---|
3 | |
---|
4 | <head> |
---|
5 | <meta http-equiv="Content-Language" content="en-us"> |
---|
6 | <title>YALMIP Example : Advanced programming</title> |
---|
7 | <meta http-equiv="Content-Type" content="text/html; charset=windows-1251"> |
---|
8 | <meta content="Microsoft FrontPage 6.0" name="GENERATOR"> |
---|
9 | <meta name="ProgId" content="FrontPage.Editor.Document"> |
---|
10 | <link href="yalmip.css" type="text/css" rel="stylesheet"> |
---|
11 | <base target="_self"> |
---|
12 | </head> |
---|
13 | |
---|
14 | <body leftMargin="0" topMargin="0"> |
---|
15 | |
---|
16 | <div align="left"> |
---|
17 | <table border="0" cellpadding="4" cellspacing="3" style="border-collapse: collapse" bordercolor="#000000" width="100%" align="left" height="100%"> |
---|
18 | <tr> |
---|
19 | <td width="100%" align="left" height="100%" valign="top"> |
---|
20 | <h2>Advanced YALMIP programming</h2> |
---|
21 | <hr noShade SIZE="1" color="#000000"> |
---|
22 | <p>In some cases, it might be necessary to do more advanced YALMIP coding. |
---|
23 | YALMIP comes with a number of commands to facilitate this.</p> |
---|
24 | <p>Let us begin by defining a set of variables.</p> |
---|
25 | <table cellPadding="10" width="100%"> |
---|
26 | <tr> |
---|
27 | <td class="xmpcode"> |
---|
28 | <pre>x = sdpvar(2,1); |
---|
29 | y = x'*randn(2)*x;</pre> |
---|
30 | </td> |
---|
31 | </tr> |
---|
32 | </table> |
---|
33 | <p>A common task is to find variables used in an expression. To do this, there are |
---|
34 | two important commands, <code>depends.m </code>and <code>getvariables.m</code>.</p> |
---|
35 | <table cellPadding="10" width="100%"> |
---|
36 | <tr> |
---|
37 | <td class="xmpcode"> |
---|
38 | <pre>depends(x) |
---|
39 | <font color="#000000">ans = |
---|
40 | 1 2</font> |
---|
41 | |
---|
42 | depends(y) |
---|
43 | <font color="#000000"> ans = |
---|
44 | 1 2</font> |
---|
45 | |
---|
46 | getvariables(x) |
---|
47 | <font color="#000000"> ans = |
---|
48 | 1 2</font> |
---|
49 | |
---|
50 | getvariables(y) |
---|
51 | <font color="#000000"> ans = |
---|
52 | 3 4 5</font></pre> |
---|
53 | </td> |
---|
54 | </tr> |
---|
55 | </table> |
---|
56 | <p>The command <code>depends.m</code> gives the actual linear independent variables, while <code>getvariables.m</code> returns the |
---|
57 | variable indices to the so called relaxed variables used in YALMIP. When |
---|
58 | a nonlinear expression is defined in YALMIP, a new variable is introduced |
---|
59 | for every monomial term. For our quadratic variable y, we need 3 monomials.</p> |
---|
60 | <table cellPadding="10" width="100%"> |
---|
61 | <tr> |
---|
62 | <td class="xmpcode"> |
---|
63 | <pre>getvariables(x(1)^2) |
---|
64 | <font color="#000000"> ans = |
---|
65 | 3</font> |
---|
66 | |
---|
67 | getvariables(x(2)^2) |
---|
68 | <font color="#000000">ans = |
---|
69 | 4</font></pre> |
---|
70 | <pre>getvariables(x(1)*x(2)) |
---|
71 | <font color="#000000"> ans = |
---|
72 | 5</font></pre> |
---|
73 | </td> |
---|
74 | </tr> |
---|
75 | </table> |
---|
76 | <p>The commands above only give the internal identifiers. To create an |
---|
77 | <a href="reference.htm#sdpvar">sdpvar</a> variable using these variables, |
---|
78 | we use the command <code>recover.m</code></p> |
---|
79 | <table cellPadding="10" width="100%"> |
---|
80 | <tr> |
---|
81 | <td class="xmpcode"> |
---|
82 | <pre>variables_in_y = recover(depends(y)) |
---|
83 | <font color="#000000">Linear matrix variable 2x1 (full, real)</font></pre> |
---|
84 | </td> |
---|
85 | </tr> |
---|
86 | </table> |
---|
87 | <p>All <a href="reference.htm#sdpvar">sdpvar</a> |
---|
88 | objects are defined by the variable indicies obtained with <code>getvariables.m</code>, |
---|
89 | and a basis. Let us begin with scalar variables to explain the concepts. |
---|
90 | The basis with respect to all variables returned from <code>getvariables.m </code> |
---|
91 | is obtained using <code>getbase.m</code></p> |
---|
92 | <table cellPadding="10" width="100%"> |
---|
93 | <tr> |
---|
94 | <td class="xmpcode"> |
---|
95 | <pre>x = sdpvar(1,1); |
---|
96 | y = sdpvar(1,1); |
---|
97 | z = pi+2*x+3*y; |
---|
98 | full(getbase(z)) |
---|
99 | <font color="#000000"> |
---|
100 | ans = |
---|
101 | |
---|
102 | 3.1416 2.0000 3.0000</font></pre> |
---|
103 | </td> |
---|
104 | </tr> |
---|
105 | </table> |
---|
106 | <p>A basis with respect to a particular variable is obtained using <code>getbasematrix.m.</code></p> |
---|
107 | <table cellPadding="10" width="100%"> |
---|
108 | <tr> |
---|
109 | <td class="xmpcode"> |
---|
110 | <pre>full(getbasematrix(z,getvariables(x))) |
---|
111 | <font color="#000000"> |
---|
112 | ans = |
---|
113 | |
---|
114 | 2 |
---|
115 | </font> |
---|
116 | full(getbasematrix(z,0)) |
---|
117 | <font color="#000000"> |
---|
118 | ans = |
---|
119 | |
---|
120 | 3.1416</font> |
---|
121 | |
---|
122 | full(getbasematrix(z,123456789)) |
---|
123 | <font color="#000000"> |
---|
124 | ans = |
---|
125 | |
---|
126 | 0</font></pre> |
---|
127 | </td> |
---|
128 | </tr> |
---|
129 | </table> |
---|
130 | <p>Matrix variables are defined in the same way, except that all basis |
---|
131 | matrices are stored in a vectorized format.</p> |
---|
132 | <table cellPadding="10" width="100%"> |
---|
133 | <tr> |
---|
134 | <td class="xmpcode"> |
---|
135 | <pre>x = pi*eye(2)+5*sdpvar(2,2); |
---|
136 | base = full(getbase(x)) |
---|
137 | <font color="#000000"> base = |
---|
138 | |
---|
139 | 3.1416 5.0000 0 0 |
---|
140 | 0 0 5.0000 0 |
---|
141 | 0 0 5.0000 0 |
---|
142 | 3.1416 0 0 5.0000</font> |
---|
143 | |
---|
144 | reshape(base(:,1),2,2) |
---|
145 | <font color="#000000"> ans = |
---|
146 | |
---|
147 | 3.1416 0 |
---|
148 | 0 3.1416</font></pre> |
---|
149 | </td> |
---|
150 | </tr> |
---|
151 | </table> |
---|
152 | <p>The command <a href="reference.htm#is">is</a> can be useful to select |
---|
153 | parts of <a href="reference.htm#set">set</a> objects</p> |
---|
154 | <table cellPadding="10" width="100%"> |
---|
155 | <tr> |
---|
156 | <td class="xmpcode"> |
---|
157 | <pre>F = set(x>0) + lmi(x*x' > 0); |
---|
158 | F_linear = F(find(is(F,'linear')))</pre> |
---|
159 | </td> |
---|
160 | </tr> |
---|
161 | </table> |
---|
162 | <p><code>replace.m</code> can be useful in some situations to replace an |
---|
163 | <a href="reference.htm#sdpvar">sdpvar</a> inside an |
---|
164 | <a href="reference.htm#sdpvar">sdpvar</a> or <a href="reference.htm#set"> |
---|
165 | set</a>.</p> |
---|
166 | <table cellPadding="10" width="100%"> |
---|
167 | <tr> |
---|
168 | <td class="xmpcode"> |
---|
169 | <pre>F0 = replace(F,x(1),0); |
---|
170 | y0 = replace(y,x(1),0);</pre> |
---|
171 | </td> |
---|
172 | </tr> |
---|
173 | </table> |
---|
174 | <p>The terms in a <a href="reference.htm#set">set</a> object can be |
---|
175 | extracted and converted to <a href="reference.htm#sdpvar">sdpvar</a> |
---|
176 | objects.</p> |
---|
177 | <table cellPadding="10" width="100%"> |
---|
178 | <tr> |
---|
179 | <td class="xmpcode"> |
---|
180 | <pre>xtimesx = sdpvar(F(2));</pre> |
---|
181 | </td> |
---|
182 | </tr> |
---|
183 | </table> |
---|
184 | <p>More to come but do not hesitate to ask.</td> |
---|
185 | </tr> |
---|
186 | </table> |
---|
187 | </div> |
---|
188 | |
---|
189 | </body> |
---|
190 | |
---|
191 | </html> |
---|