You can configure the graph axes to display log, log2, ln, probability, probit, reciprocal and offset reciprocal scales, as well as linear scales.
In addition, you can combine different axis scales in a single graph by linking to a second layer or by creating a nonlinear scale on a second axis in the same layer.
Standard linear scale: X'=X.
Base 10 logarithmic scale: X'=log(X).
Represents the inverse of a cumulative Gaussian distribution. Plotting a cumulative Gaussian distribution produces a sigmoidally-shaped curve. This curve, when displayed on a probability scale, appears as a straight line. Since probabilities are expressed as percentages, all values must fall between 0 and 100. The probability scale range is 0.0001 to 99.999.
Like the probability scale, a sigmoidally-shaped curve plots as a straight line. In this case, however, the scale is linear, and the increment between tick marks is exactly one standard deviation. The value "5" on the scale shows the mean, or 50% probability. "6" is one standard deviation away, etc.
Reciprocal scale, where X'=1/X.
Offset reciprocal scale, where X'=1/(X+offset). Offset is defined as 273.14, where 273.14 is the absolute temperature for 0° C.
Logit=ln(Y/(100-Y)). As with the probability and probit scales, a sigmoidally-shaped curve plots as a straight line.
Natural log scale (base e logarithmic scale).
Base 2 logarithmic scale.
A note about log10 scales:*
Starting with Origin 7, if the log scale range is within one decade, the ticks and grids will be linear.
A LabTalk system variable @TL determines whether to use linear tick marks by the following relation:
Thus, to support linear tick locations for two decades, for example, set this variable to 14 by entering the following at the command line:
The default value for @TL is 10.
The offset reciprocal scale type is defined as X'=1/(X+273.14), where 273.14 is the absolute temperature corresponding to zero degrees Celsius.
We can create a second layer and add a custom link between the first layer's offset reciprocal scale and the second layer's linear scale and use the resulting graph window to plot temperature data on the two scales.
Using a LabTalk Script command, you can create a nonlinear scale on a top X or a right Y axis in a 2D graph, based on values of the bottom X or left Y axis. This is a simple procedure, and does not require creation of aUsing a LabTalk Script command, you can create a nonlinear scale on a top X or a right Y axis in a 2D graph, based on values of the bottom X or left Y axis. This is a simple procedure, and does not require creation of a second graph layer. Tick marks on the second axis are located opposite tick marks on the primary axis.
For example, in this graph the bottom X axis scale reads in some quantity X, while the top axis scale is in units of 1/X.
In order to create this graph, you would do the following:
This computes and places labels on the top X axis, at positions corresponding to the tick marks and labels of the bottom X axis, using the formula specified between the quotation marks. This formula can be any function of X that is calculable using LabTalk.
Note: Similar to the top X axis in this example, you can turn on right Y axis and create labels for that axis too. However, you should not refer to "Y" in your formula; rather you would still use "x," as in:
In the last procedure, we used a formula to create a nonlinear scale on a second x axis. One drawback for some users will be that the scale values displayed on the second axis are not nicely-rounded numbers. If you (1) do not need ticks on the second axis to align with ticks on the primary axis and (2) you want control over the scale values that display on the second axis, there is an alternate procedure.
Note that in this graph, the values on the nonlinear scale on the top X axis are derived from the reciprocal of the values on the linear scale on the bottom X axis.
The following steps describe how to create such a graph:
Note: To set up a similar relationship for a second Y axis, you would need to alter your command slightly:
layer2.y.ticksbydata$=<dataset containing tick mark positions>