From: André Wobst <wobsta@us...>  20050919 05:44:37

Update of /cvsroot/pyx/pyx/examples/graphs In directory sc8prcvs1.sourceforge.net:/tmp/cvsserv26750/examples/graphs Added Files: lissajous.txt Log Message: some first example descriptions  NEW FILE: lissajous.txt  The example shows how to use a `paramfunction` as a data source for a graph. `paramfunction` defines a parameter variable name given as a string in the first argument, a minimal and a maximal value given as the second and third argument of `paramfunction` and an expression string as forth argument. This expression string assigns a tuple of expressions to a tuple of data names. As usual the styles will decide what those data names are responsible for. Like for the function, you can also access external data and functions in your expression. Suppose we want to provide the data for this example by the following python function: def lissajousdata(k): return sin(2*k), cos(3*k) Then we would need to modify the plot command to: g.plot(graph.data.paramfunction("t", min, max, "x, y = f(t)", context={"f": lissajousdata})) Note, that `t` and `f` in the string expression stand for `k` and `lissajousdata` respectively. You can also use the same names in the expression if you like. When you try this, you'll notice, that you need to add `sin` and `cos` from the list of imported symbols from the math module. While this is expected, it is more interesting to notice, that you ''didn't'' need to import those functions before. The reason is, that the string expressions are evaluated in a certain context. This context doesn't only contain the symbols given in the `context` keyword argument, but additionally some standard mathematical functions (and constants) including `sin` and `cos`. See the reference manual for a complete list. For the common case of plotting a parametric function providing data for `x` and `y` by an external function, there is a special `paramfunctionxy` which directly calls the external function instead of using a context in a string expression. By a `lambda` expression we could provide the data inline, which results in: g.plot(graph.data.paramfunctionxy(lambda k: (sin(2*k), cos(3*k)), 0, 2*pi)) Note that the external expression is provided in the first argument here. Since the system just need to call the function, the parameter variable name doesn't need to be provided to `paramfunctionxy`. 