A CLOSER LOOK ATGRAPH WINDOWS
We now take a look at the menus and tools available with a graph or plot window . The toolbar is as shown below, and each tool is numbered for reference.
| 1 |
2 |
3 |
4 |
5 |
6 |
7 |
8 |
9 |
10 |
11 |
12 |
13 |
14 |


Operating on charts in place:
Call up a chart in the Graph Wizard. Click on the data series itself to “select” it, which means if there are 2 charts, or more, you are selecting one vs. another. Clicking on the chart page is not specific enough, you actually have to select the data series. When you have selected it, the chart will be highlighted. If the data series charted is too close together to select one vs. another, then click on the "MAGNIFYING GLASS" icon to do so.
![]() |
![]() |
|
the chart “unselected”
|
the chart “selected”
|
Next, click on the rectangle icon, which will enable you to highlight the time period you want to examine.
Then go and highlight the data, in much the same way as data is selected and highlighted in the automatic mode of the parabolic prediction process. (See Parabolic Prediction Introduction).
With the data highlighted, go to the
icon in the upper right, you
will then be presented with the following palate, which enables you to select
what you wish to do with that specific data.
Note that on this palate we have selected all of the choices. We have done this so you can see the type of parameters you will need to specify, depending on which choices you select. Obviously if you select all of the choices, it’s going to be a rather confusing chart. Once you have made your selections, press Begin Construction.
You are then presented with a chart with the regressed fit or smoother applied. But that’s not all! The extra line (or lines) you have created is not just a picture on the chart, but an actual data series, with its formula. Like any data series, it can be manipulated, extended or saved. To see it, just click on the view data tab (at the bottom), which will show you the dataset “on the flip side of the chart”. Do whatever you want with the dataset, and then click on the plot tab to get back to a chart. Note that the data and graph tabs are alternates to each other. Anything you have done to alter the dataset will be taken into consideration and plotted when you get back to the graph.
With this Operating on Charts in Place feature, you can visually select a present period and extend the fit forward, essentially making a current forecast. You can also pick a period out of the past, fit the data, make a forecast, and see how well your forecast performed. That is, you are not back fitting a system, but being intellectually true.
Choosing parameters:
If the data series you have selected on the chart is a bar, recognize that it is 3 or 4 datasets (i.e. the high, low, close and possibly the open). Thus the smoothing choices will create an equal number of smoothing lines.
Regression Fits (either linear or polynomial) of “all Data together” mean that the resulting output is a single line. For a bar chart, the high, low, close, and open (if available) are all considered to achieve the “fit”.
Moving Average: Should be a single integer.
Exponential Moving Average: The N value (see the manual under expave).
Weighted Moving Average: See the manual under wtave.
Cyclic Approximation: Should be a single integer.
Degree of Polynomial: Should be an integer. 1 = linear, 2 = polynomial,
3 = cubic.
The File menu drops down as shown below. The "Save" option works exactly as the corresponding tool. The second option is "Save As" which prompts for a name or a new name, then saves the current Dataset under that name, complete with the current plotting format. The fourth option saves the plot layout under a name you might choose. This layout can be recalled and applied to any similar Dataset with the plotting wizard. Finally, the "Print" option functions exactly like the corresponding tool.
Next we see the Edit menu in drop down form.

The "Cut", "Copy", and "Paste" selections are
identical with the same tools on the toolbar. The "Clear" selection
simply eliminates the selected graphs. These The "Extend" selection
will extend in place any of the selected graphs that are extendable (that is
they were created with, and retain, mathematical formulas). If no Datasets are
selected in the pane of the current plot window, all options other than "Paste"
will be inactive. The "Paste" option is active whenever there is a
graph on the plot clipboard.
Finally, the Help menu displays this section on tools and menus associated with
the Dataset window.
![]()
Another choice you will see on the plot menu is “Save Selection in New Window” This is a great feature that behaves like a “Copy and Paste” operation, but for charts. Suppose you want to take part of your present chart and work on it separately, just select the data copied by clicking on it with the mouse (it will become bold), and then clicking on the “Save Selection in New Window” line under the “File” menu. That operation will result in the entire dataset of the selected data being created in another chart. The chart formats of the original chart will be preserved in the new chart, and the flip side of that chart will have a dataset of data entries. If you take a look at the command window, you will also note that the appropriate entry will have been entered for you as a command.
Now suppose you only want part of the dataset. In that case, select the dataset
by clicking on it, and then click on the Rectangle icon above. You will then
be able to select the part of the dataset you want, and click again when you
have finished. Then click on the menu item “Save Selection in New Window”.
If there are more than one datasets that you are interested in copying over,
select the first, then hold down the Shift key and select the second, etc.
Should you not be able to separate the datasets visually for selection (i.e.
they are too close together), then go to the Magnifying Glass icon, which
will enable you to select them on a menu.
The “Save Selection in New Window” feature has the tremendous advantage of
allowing you to tinker with some of the data without messing up your original
chart, which of course, is still where you have left it.
DECORATING CHARTS
If you right-click on any chart page you will get the decoration menu displayed
below. You may also access it by selecting “Add/Edit Decorations” from the
“Edit” menu.

Essentially all you have to do is select the type of decoration, then point and click on the chart. The decoration will then appear on the chart. For example, in the case of arrows, select the size, color and direction, and then click on the chart where you want the arrows located.
With regard to text, select the color, font (including size) and direction,
and then click on the chart location. You will get a window allowing you to
enter the text. There are two tricks to this process. First, remember to click
on the location accurately before you type in the text. Second, note that
the angle is measured from the normal line (which is 0 degrees). Ninety degrees
would point straight up, and 270° would point straight down. Mostly you
are going to use 90°, 45° (sloping up) and 315° (sloping down).
The paint tool allows you to paint individual bars of a bar chart, or datapoints of a line chart, in any chosen color. This will help to emphasize certain chart occurances.
The Show Logical button allows you to select any logical (True-False)
expression, involving any dataset or sets, and then have the target data painted
exactly on the days when the expression was true. Put up a chart, say EURO
LAST 1000. Then go to EDIT, ADD/EDIT DECORATIONS, and pick the color of your
choice, then SHOW LOGICAL, and select the paintbrush. You will be presented
with an input window. Let's enter, (5 CHANGE CL IRX) GT 0, and click on the
button below.

Your chart will now be painted red where the condition is true. That is, you can now see the effect on the Euro of a 5-day positive change in T-bill rates. This example demonstrates that the logical condition can be any logical condition, not just one dealing with the charted dataset.
