| Excerpt |
|---|
This example demonstrates functions for comparing the relative values of two functions. |
Functions:
| D generate list excerpts |
|---|
| pages | LESSTHANEQUAL Function,LESSTHAN Function,GREATERTHANEQUAL Function, GREATERTHAN Function |
|---|
|
In the town of Circleville, citizens are allowed to maintain a single crop circle in their backyard, as long as it confirms to the town regulations. Below is some data on the size of crop circles in town, with a separate entry for each home. Limits are displayed in the adjacent columns, with the inclusive columns indicating whether the minimum or maximum values are inclusive.
| Tip |
|---|
Tip: As part of this exercise, you can see how to you can extend your recipe to perform some simple financial analysis of the data. |
Source:
| Location | Radius_ft | minRadius_ft | minInclusive | maxRadius_ft | maxInclusive |
|---|
| House1 | 55.5 | 10 | Y | 25 | N |
| House2 | 12 | 10 | Y | 25 | N |
| House3 | 14.25 | 10 | Y | 25 | N |
| House4 | 3.5 | 10 | Y | 25 | N |
| House5 | 27 | 10 | Y | 25 | N |
Transformation:
After the data is loaded into the Transformer page, you can begin comparing column values:
| D trans |
|---|
| RawWrangle | true |
|---|
| p03Value | 'tooSmall' |
|---|
| Type | step |
|---|
| WrangleText | derive type:single value: LESSTHANEQUAL(Radius_ft,minRadius_ft) as:'tooSmall' |
|---|
| p01Name | Formula type |
|---|
| p01Value | Single row formula |
|---|
| p02Name | Formula |
|---|
| p02Value | LESSTHANEQUAL(Radius_ft,minRadius_ft) |
|---|
| p03Name | New column name |
|---|
| SearchTerm | New formula |
|---|
|
While accurate, the above transform does not account for the minInclusive value, which may be changed as part of your steps. Instead, you can delete the previous transform and use the following, which factors in the other column:
| D trans |
|---|
| RawWrangle | true |
|---|
| p03Value | 'tooSmall' |
|---|
| Type | step |
|---|
| WrangleText | derive type:single value: IF(minInclusive == 'Y',LESSTHANEQUAL(Radius_ft,minRadius_ft),LESSTHAN(Radius_ft,minRadius_ft)) as:'tooSmall' |
|---|
| p01Name | Formula type |
|---|
| p01Value | Single row formula |
|---|
| p02Name | Formula |
|---|
| p02Value | IF(minInclusive == 'Y',LESSTHANEQUAL(Radius_ft,minRadius_ft),LESSTHAN(Radius_ft,minRadius_ft)) |
|---|
| p03Name | New column name |
|---|
| SearchTerm | New formula |
|---|
|
In this case, the IF function tests whether the minimum value is inclusive (values of 10 are allowed). If so, the LESSTHANEQUAL function is applied. Otherwise, the LESSTHAN function is applied. For the maximum limit, the following step applies:
| D trans |
|---|
| RawWrangle | true |
|---|
| p03Value | 'tooBig' |
|---|
| Type | step |
|---|
| WrangleText | derive type:single value: IF(maxInclusive == 'Y', GREATERTHANEQUAL(Radius_ft,maxRadius_ft),GREATERTHAN(Radius_ft,maxRadius_ft)) as:'tooBig' |
|---|
| p01Name | Formula type |
|---|
| p01Value | Single row formula |
|---|
| p02Name | Formula |
|---|
| p02Value | IF(maxInclusive == 'Y', GREATERTHANEQUAL(Radius_ft,maxRadius_ft),GREATERTHAN(Radius_ft,maxRadius_ft)) |
|---|
| p03Name | New column name |
|---|
| SearchTerm | New formula |
|---|
|
Now, you can do some analysis of this data. First, you can insert a column containing the amount of the fine per foot above the maximum or below the minimum. Before the first derive command, insert the following, which is the fine ($15.00) for each foot above or below the limits:
| D trans |
|---|
| RawWrangle | true |
|---|
| p03Value | 'fineDollarsPerFt' |
|---|
| Type | step |
|---|
| WrangleText | derive type:single value: 15 as:'fineDollarsPerFt' |
|---|
| p01Name | Formula type |
|---|
| p01Value | Single row formula |
|---|
| p02Name | Formula |
|---|
| p02Value | 15 |
|---|
| p03Name | New column name |
|---|
| SearchTerm | New formula |
|---|
|
At the end of the recipe, add the following new line, which calculates the fine for crop circles that are too small:
| D trans |
|---|
| RawWrangle | true |
|---|
| p03Value | 'fine_Dollars' |
|---|
| Type | step |
|---|
| WrangleText | derive type:single value: IF(tooSmall == 'true', (minRadius_ft - Radius_ft) * fineDollarsPerFt, 0.0) as: 'fine_Dollars' |
|---|
| p01Name | Formula type |
|---|
| p01Value | Single row formula |
|---|
| p02Name | Formula |
|---|
| p02Value | IF(tooSmall == 'true', (minRadius_ft - Radius_ft) * fineDollarsPerFt, 0.0) |
|---|
| p03Name | New column name |
|---|
| SearchTerm | New formula |
|---|
|
The above captures the too-small violations. To also capture the too-big violations, change the above to the following:
| D trans |
|---|
| RawWrangle | true |
|---|
| p03Value | 'fine_Dollars' |
|---|
| Type | step |
|---|
| WrangleText | derive type:single value: IF(tooSmall == 'true', (minRadius_ft - Radius_ft) * fineDollarsPerFt, if(tooBig == 'true', (Radius_ft - maxRadius_ft) * fineDollarsPerFt, '0.0')) as: 'fine_Dollars' |
|---|
| p01Name | Formula type |
|---|
| p01Value | Single row formula |
|---|
| p02Name | Formula |
|---|
| p02Value | IF(tooSmall == 'true', (minRadius_ft - Radius_ft) * fineDollarsPerFt, if(tooBig == 'true', (Radius_ft - maxRadius_ft) * fineDollarsPerFt, '0.0')) |
|---|
| p03Name | New column name |
|---|
| SearchTerm | New formula |
|---|
|
In place of the original "false" expression (0.0), the above adds the test for the too-big values, so that all fines are included in a single column. You can reformat the fine_Dollars column to be in dollar format:
| D trans |
|---|
| RawWrangle | true |
|---|
| Type | step |
|---|
| WrangleText | set col: fine_Dollars value: NUMFORMAT(fine_Dollars, '$###.00') |
|---|
| p01Name | Columns |
|---|
| p01Value | fine_Dollars |
|---|
| p02Name | Formula |
|---|
| p02Value | NUMFORMAT(fine_Dollars, '$###.00') |
|---|
| SearchTerm | Edit column with formula |
|---|
|
Results:
After you delete the columns used in the calculation and move the remaining ones, you should end up with a dataset similar to the following:
| Location | fineDollarsPerFt | Radius_ft | minRadius_ft | minInclusive | maxRadius_ft | maxInclusive | fineDollars |
|---|
| House1 | 15 | 55.5 | 10 | Y | 25 | N | $457.50 |
| House2 | 15 | 12 | 10 | Y | 25 | N | $0.00 |
| House3 | 15 | 14.25 | 10 | Y | 25 | N | $0.00 |
| House4 | 15 | 3.5 | 10 | Y | 25 | N | $97.50 |
| House5 | 15 | 27 | 10 | Y | 25 | N | $30.00 |
Now that you have created all of the computations for generating these values, you can change values for minRadius_ft , maxRadius_ft , and fineDollarsPerFt to analyze the resulting fine revenue. Before or after the transform where you set the value for fineDollarsPerFt, you can insert something like the following:
| D trans |
|---|
| RawWrangle | true |
|---|
| Type | step |
|---|
| WrangleText | set col: minRadius_ft value:'12.5' |
|---|
| p01Name | Columns |
|---|
| p01Value | minRadius_ft |
|---|
| p02Name | Formula |
|---|
| p02Value | '12.5' |
|---|
| SearchTerm | Edit column with formula |
|---|
|
After the step is added, select the last line in the recipe. Then, you can see how the values in the fineDollars column have been updated.
| D s also |
|---|
| label | example_comparison_functions2 |
|---|
|
