Troubleshooting

How do I add variables or data to an existing dataset?
You can select the functions to import new data files (CSV or XLS) to a dataset, but this adds files not previously imported to previous datasets to a new dataset (see“Import Data Files”).
  • If you have data files from different data sources (e.g., in files), import these files in the same import step so that they are available within a common dataset.
  • If data is imported independently, Data Explorer creates a new independent dataset. Data from two independent datasets may be plotted together by selecting the desired variables from multiple datasets.
When you export data, data may be selected from multiple datasets through the Export page. You will probably want to align data via common Time Merge entries (e.g., Start, End, and Interval Times) and, if used, relevant overlapping filters.
How do I change a variable name?
Right-click the variable and click
Rename
.
How can I have my plots indicate both name and additional dataset parameters?
In the overview page, click [Configure Properties]. Additional properties can be selected/starred for presentation when your mouse hovers over values in a plot.
There is a date/time variable that the importer cannot interpret.
When importing your dataset (see“Import Data Files”):
  1. Under
    Specify File Structure
    , verify your file locale and time zone provide appropriate information for this data(see Figure 4 )
  2. Under
    Format Column Data
    , check your date/time format settings and map it to your actual date time format (Figure 5).
  3. If the date/time cannot be interpreted by the importer, in the importer set the variable Type to String
  4. After you finish formatting the file and create a dataset from it, use a transform to build a date/time variable from the input string. For example,
DtNew = $dtread($str($left(!dt!, 10),' ', $right(!dt!,4),' ',$left($right(!dt!, 13),8)))
  1. In the
    Column Configuration
     tab, click on the newly-generated variable, then right-click and select
    Duplicate Variable (Raw)
     to create a new raw date/time variable; then delete the transform that you builtin step 4.
  2. Change the view to
    Before Transform
     (raw data).
  3. Do a muti-select across the assigned Time Column of the variables you want to assign to the new date time variable.
  4. Press your space key (this opens a selection button for multiple selected table cells) and select DtNew (or the appropriate time column) and click [OK].
I made a dataset, and it thinks that some of my numbers are strings.
It is best to correct this problem in the import wizard before you create the dataset or redo the import with corrected assignments. There are several situations that can cause a column of raw data to be misinterpreted as strings. In
Format Column Data
 of the import wizard (Figure 5), you should check the type of every column and correct them if necessary.
An alternative method is available if it is not practical to go back to the importer. For example, if the variable flow1 is a string but should have been numeric:
  1. Use the $val transform to create a new numeric variable.
flow1_num. flow1_num = $val(!flow1!)
  1. Because flow1_num is a computed column, use the Duplicate Column (raw) from the Overview screen (manifest) on this variable to copy it into a new raw variable (flow1_raw).
  2. You no longer need flow1_num or the original flow1 (string variable), so delete them from the system manifest.
  3. You may then rename the flow1_raw variable (after deleting the original flow1 variable) to flow1.
The dataset already has a time merge, but now I have added some more variables, and they must be time merged as well. Do I have to sit through two time merges every time it evaluates the transform list?
Time merges frequently (but not always) support editing via the configuration window. If supported, you will be asked if you want to edit or create a new time merge and can select your preferred option.
If not supported (frequently on legacy datasets constructed with earlier Pavilion8 tools), you may edit your transforms, locate the time merge transform, and add additional Date/Time columns to the set of the last transform entries. Add a comma after each additional TimeColumn you want to include in the same TimeMerge with the desired settings.
Two variables have the same Date/Time column, but I want them to use different Time Merge methods.
Duplicate your Date/Time column.
In the Column Properties screen in the dataset overview, change the second set of desired variables to use the new Time Column as their Time Column.
Now create two-time merge applications using the desired Time Columns with the desired configuration.
I want a transform to make a real number, but it comes out as an integer.
Use the $real transform to force the output to be a real number. For example, for a dataset in which! real! is a real number and !igr! is an integer
!new! = $if(!real! > !igr!, !real!, !igr!)
produces an output variable that is an integer, but
!new! = $real($if(!real! > !igr!, !real!, !igr!))
produces an output variable that is a real number.
I need to find Quartile or “Decile,” limits in the data distribution.
  1. 1. Make a histogram plot of the variable (see“Histogram”).
  2. Check the Show accumulation checkbox.
  3. Set the Bin Count (right mouse click on the chart) to 500 (maximum) so that you cannot see the individual bins.
  4. Move your cursor up and down and note that both the value (mid-bin) and percentage of the total is provided. You can identify desired/closest distribution percent and note the associated presented value in the pop-up.
Provide Feedback
Have questions or feedback about this documentation? Please submit your feedback here.