Month: November 2017

TERR — Converting strings to date and time

This post explains my struggle to convert strings to Date or Time with TERR.  I recently spent so much time on this that I thought it deserved a blog post.  Here’s the story…

I was recently working on a TERR data function that calls a publicly available API and brings all the data into a table.  I used the function to parse out my row data.  In that function, I used the stringsAsFactors = FALSE argument, and as a result (the desired result), all of my data came back as strings.  This was fine because the API included column metadata with the data type.  As you can see in the script below, I planned on “sapplying” through the metadata with as.POSIXct and as.numeric.  This worked just fine in RStudio, and it also worked for the numeric columns and for the DateTime columns.  However, it did not work for Date and Time columns.  I tried different syntax, functions (as.Date didn’t work either), packages, etc to get it to work and NOTHING!  The struggle was very real.

Script convert strings to Date or Time with TERR


Finally, I Googled the right terms and came across a TIBCO knowledge base article with this information….

Spotfire data functions recognize TERR objects of class “POSIXct” as date/time information. As designed, the Spotfire/TERR data function interface for date/time information does the following:

– Converts a Spotfire value or column whose DataType is “Date”, “Time” or “DateTime” into a TERR object of class “POSIXct”.

– Converts a TERR object of class “POSIXct” into a Spotfire value or column with a DataType of “DateTime”, which can then be formatted in Spotfire to display only the date (or to display only the time) if needed.

This interface does not use any other TERR object classes (such as the “Date” class in TERR) to transfer date/time information between Spotfire and TERR.

That told me that all my effort was for naught, and it just wasn’t possible.  I contacted TIBCO just to make sure there wasn’t some other solution out there that the article was not addressing.  In the end, I just used a transformation on the Date and Time columns to change the data type.  I hope that you, dear Reader, find this post before you spend hours on the same small problem.  I did put in an enhancement request.  Fingers crossed.  Please let me know if you have a better method!



Guest Spotfire blogger residing in Whitefish, MT.  Working for SM Energy’s Advanced Analytics and Emerging Technology team!

Part 5 – Data Function Properties

This is the fifth part of a series on Spotfire Properties.  In previous posts, I discussed Document PropertiesData Table PropertiesColumn Properties, and Data Connection Properties.  This week we’ll take a look at data function properties.  Next week, the series will conclude with Visualization Properties.  Data function properties is a bit of a beefier subject because all data functions can be created out of the data function properties menu.

What is a Data Function?

Since I am writing this for the new Spotfire user, let’s start with the basics. What is a data function, and why would you need one? A data function is TERR code written to perform a specific task within Spotfire. What kind of tasks you ask? I’ll come back to that in just a sec. First, I want to define TERR. TERR stands for TIBCO Enterprise Runtime for R. TERR is the Spotfire version of open source R. R is a programming language for statistical computing and graphics. The R language is widely used among statisticians and data miners for developing statistical software and data analysis (Wiki). 
As it relates to Spotfire, TERR expands upon the functionality of the application. For example, is about to post a template that uses TERR to identify duplicate records (link coming soon).  You can use TERR to manipulate data, pass data thru to new tables, and expand upon the statistical and predictive modeling tools that come with the application. For example, Spotfire doesn’t have a Random Forest model, but you can build one with TERR.  The only limits are really your own programming skills. 
Now, I’ve already written a blog post on TERR basics, which includes an explanation of all the TERR screens.  Most of what is in that post I would have also covered here.  Therefore, I am going to use this post to expand on the different options in the screens, rather than walking through a single use case.  Lastly, also has a few templates such as this template and this template to get you started.    

Data Function Basics

Let’s review the basics of Data Functions.  Creating data functions is a 3 step process.
  1. Create the script
  2. Create the parameters
  3. Run the script to map the parameters to the data in the DXP
Key concepts to be aware of:
  • The script is the “meat” of the data function. Within the script, you’ll find at least one input and one output parameter.  The simplest R script I’ve ever written is output <- input.  Input is the input parameter, and…yeah, you can finish that sentence I bet.  
  • TERR (and R) are object-oriented languages, which means programmers can create objects within the code, assign values to the objects and then reference the object down the road rather than all the values.  This makes programming easier. In the example above, input and output are both objects.
  • Input and output parameters tell Spotfire what type of object to work with.  The object could be a table, column, document property or another object. 
  • Running the script triggers the dialogs where you will map the parameters to the actual data in the DXP.
  • Data functions can be connected to marking and filtering.  For example, you can pass the results of marking or filtering to a new table.
  • Users may create data functions from scratch in Spotfire, or users may import data functions from the Spotfire library or another file.
  • By default, data functions embed within the analysis. However, users have the ability to save them to the library for reuse or sharing.
With that said, here is a very simple script that I’ll reference throughout the post.  This single line of R code that will check for duplicates in a data set using two identifiers that define the granularity of the table. The output is a column called duplicate.

duplicate <- duplicated(data.frame(key1, key2))

duplicated and data.frame are TERR (or R) functions
duplicate, key1, and key2 are objects in the script
key1 and key2 are input parameters
duplicate is the output parameter

Data Function Properties Main Screen

This dialog lists all data functions in the DXP.  From this screen you can …
  • Create new data functions
  • Edit existing data functions and their parameters
  • Refresh data functions
  • Delete data functions
  • Save data functions to the library
  • Export a data function
Data Function Properties
Related to the buttons above:
  • The term “Register New” can be a bit confusing to new users.  This really means create a new data function.  In the process, you’ll have the option to save it in the library or register it.  
  • Clicking the Edit Script button will let you modify the script or the input and output parameters.
  • Clicking the Edit Parameters button allows you to change the mapping of data from the parameters to the DXP content.

Script & Parameters

These three dialogs define the script, input parameters, and output parameters.  Because the TERR Basics post covers them in detail, I want to focus on the different options available to the input and output parameters.  
Data Function PropertiesData Function PropertiesData Function Properties

Input and Output Parameters

When creating input and output parameters, there are three options — table, value, and column.  ‘
  • If your input or output is an entire table, choose Table.  I use this option when I am simply passing a limited data set from my original table to a new table.
  • If your input or output is a single column, choose Column.  The script shown above for identifying duplicates uses a Column output.  The data function creates a column called “duplicate”.
  • If your input is a hard-coded value or a document property, choose Value.

Data Function Properties

Data Function Properties

Run the Data Function

After you have entered the script, input parameters, and output parameters, the next step is clicking the Run button.  If Spotfire asks if you want to save the data function to the library, you can say no.  It will not impact your DXP.  This is simply to give the option to save the data function to the library so others may access it.  As an administrator, I ask users NOT to do this because it clutters up the library.  It is also hard to know what a given data function is for or if it even works.

Input Parameters

Anyway, this is the step in the process where you map the parameters to the content of the DXP.  Let’s tackle the inputs first.  I have intentionally added two unnecessary parameters to demonstrate that the options for input handlers depend on the type of input parameter.  Each input parameter type has different options.

  • For Column type, there are three options — Column, Expression, and None.  The most common input handler is Column, which I have used in data functions that manipulate or calculate based on a specific column of data.
  • For Value type, there are six options — Value, Document property, Data table property, Column property, Expression, and None.  I most frequently use Document property.
  • For Table type, there are three options — Columns, Expression, and None.  You can tell Spotfire to work with a subset of the columns in the table by using the Select Columns button. Alternatively, typing “*” in “Search expression” will use all columns in a table.  It’s not visible in the screenshot shown, but just below the “Search expression” section, you will also find options to connect the contents of the table to marking or filtering.  This is explained in the TERR Basics post.

I do want to note that I have never used the None option in either input or output handlers.  If someone has, please tell me about it in Comments.

Data Function Properties

Data Function Properties

Data Function Properties


Output Parameters

Now, for outputs, it is also true that the options presented differ depending on the parameter type.  As you can see, Column, Value, and Table all have different options.

  • The Column and Table Type have the same four options — Data table, Columns, Rows, and None.  Use Data table if you are creating an entirely new table.  Set the type to Columns if the output is a column that should be added to another table.  Use Rows if you are adding rows to a table.
  • In Value Type, there are six options — Data table, Columns, Rows, Document property, Data table property, Column property, and None.  The same advice is true of outputs here as for inputs.

Data Function Properties Data Function PropertiesData Function Properties


As I was writing this, I realized that if I were creating a data function that output rows, I’m not sure which type I would use.  The options for adding rows are part of both the Column and Table Type.  Setting up a Column type to insert rows seems counter-intuitive.  I just haven’t had to write this type of data function yet.  If you know, please Comment!

Hopefully, explaining some of the common uses of the different types of input and output parameters will help you better understand TERR function and how to convert R code to TERR.  Thanks!






Guest Spotfire blogger residing in Whitefish, MT.  Working for SM Energy’s Advanced Analytics and Emerging Technology team!

Part 4 – Data Connection Properties

This is the fourth part of a series on Spotfire Properties.  In previous posts, I discussed Document Properties, Data Table Properties, and Column Properties.  This week we’ll take a look at Data Connection Properties.

Data Connections

When I first started working with Spotfire, I didn’t know what a data connection was.  Thus, I didn’t understand the dialog at all or what it did.  So, let’s first talk about what a data connection is.

When the application opens, there are four options for adding data tables, as shown below.  If you added a data table using the last option, the Data Connection Properties dialog will be relevant.  If you used one of the first two options, you do not have any data connections in the DXP, and you don’t need to worry at all about the Data Connections dialog.

Data Connection

When you add a data table with a data connection, you are making a direct connection to the data source (such as SQL databases).  This type of connection will usually require Windows authentication or some type of username and password.  Many companies do not allow Spotfire users to make this type of connection because it hits the database directly, which can impact resources and other processes.  This is a meaty subject, but for now, just understand that connecting with a data connection to a database is different from connecting to a database with an information link.

Data Connection Properties

The Data Connection Properties dialog opens up to the screen shown below where you have the option to Rename the connection, go to Settings, or Refresh the data.  I highly recommend coming into this screen after adding a data connection to rename the connection.  Why?  Well, when you import data from a data connection, most users name the table, but they don’t know they can also name the connection.  When the DXP has many connections, they can get confusing quickly, and so it’s a best practice to rename them.  In this case, the delete button is grayed out because the DXP only has one table from one connection, so deleting it isn’t an option.

Data Connection Properties


If you click on the Settings button in Data Connection Properties, a new dialog box appears with three tabs.  By default, the data connection is embedded in the analysis, but you can also choose to save the connection in the library for other users.  This dialog also has an option to replace the data connection.

Settings – General

From within the General tab, you can see the views and tables associated with the data connection.  There are two actions you can take — Edit the data connection and Refresh the Schema.  Prior to version 7.11, clicking on the Edit button was the only way to edit the data coming in.  However, this type of editing is now available in the Data Panel as well.  Refreshing the Schema will simply ping the data connection to show you any changes to the schema.  This could be new tables, new views, or other changes to the database.

Data Connection Properties General

Settings – Data Source

The data source tab shows all of the information about the data connection itself.

Data Connection Properties Data Source

Settings – Login

Clicking on the Settings button will bring up another dialog with the details of the connection and two more tabs.  You can edit the data connection in the Login tab by clicking the Edit button.  For example, if you were working with a data table in a “dev” or “uat” instance, you can change the connection to the “prod” instance.  As long as everything is the same, the tables are easily swapped out without any loss of inserted columns, transformations, or calculations.

Data Connection Properties Credentials

Settings – Credentials

The Credentials tab is very important, perhaps the most important dialog for Data Connections.  There are three ways credentials may be retained in Spotfire.

  1. No, do not save any credentials — This means that each time a new user attempts to open the DXP, they will need to enter their own database credentials in order to update the table.  This can be problematic depending on how free your company is about handing out database credentials.
  2. No, but save credentials profile — This will save a credentials profile, which consists of a profile name, a username, and a password (only the profile name is saved with the connection data source).
  3. Yes, save credentials… — This can be risky, but I have used this option before when had to use a data connection and had many users without direct database access.  In this case, we setmup a generic account (rather than an account for each user).  As the administrator, I set up the DXP, entered the username and password.  Spotfire saved it, and the users didn’t need to know either username or password, and the file worked.

The third option does not appear in the screenshot below because I opened the data connection with Windows authentication, rather than SQL Authentication.  You would see this option when it is relevant.

Data Connection Properties Settings

Settings – Cache Settings

Finally, the Cache Settings tab specifies when data should refresh.  How you configure these settings will vary depending on…

  • The frequency of data updates
  • The number of users using the direct connection
  • The volume of data being loaded

Spotfire provides three options.

  1. No, always get fresh data from the external source.  Use this option if you do not want to cache data from the data connection.  This may put a very high load on the database.
  2. Yes, but let the cached data expire.  Use this option to cache data but refresh if the cached data is older than a specified limit.
  3. Yes, but let the cached data expire every (specify interval).  Use this option to cache data but refresh at a specified time or interval.

Data Connection Properties Settings

Hopefully, you have a better understanding of what Data Connections are and what options you have to configure them.  The next post will look at Data Function Properties.

Guest Spotfire blogger residing in Whitefish, MT.  Working for SM Energy’s Advanced Analytics and Emerging Technology team!

Part 3 – Spotfire Properties – Column Properties

This is the third part of a series on Spotfire Properties.  In previous posts, I discussed Document Properties and Data Table Properties.  This week we’ll take a look at Column Properties.

Column Properties

As the name indicates, the Column Properties dialog deals with properties for columns in a data table.  Now, that explanation seems clear and simple, but when making the switch from Excel to Spotfire, there are a few things that get lost in translation.  That’s okay.  We’ll work thru it in this post.

Column Properties contains 5 tabs.  No matter which tab you click on, the data table drop down and the list of column names will always be visible.  You will also always see the Insert and Delete buttons.  The Insert button allows you to insert a calculated column, a binned column, or a hierarchy column.  This might be preferable to using the Insert menu because you can insert more than one calculated column at a time without exiting the dialog.

Additional features of this main dialog include:

  • Type to search — This is almost always faster than scrolling, and you can use wildcards (“*”).  For example, if I want to search for any column of data with the word date, I would search “*date*”.
  • Click to sort — Users may also click on the column headers to sort.  I frequently use this when I create new information links, and I need to check for Currency data types.  I click the Column Type header and can quickly see if I have any Currency columns.
  • Metadata — There is a ton of information about each column.  See below for more detail.

Column Properties Header


Meta Data

As you can see from the screenshot, there is a ton of information about each column.  This is too much to cover in one post, so I’ll just go over the most important information.

  • Column Type — There are 9 column types — None, Imported, Calculated, Binned, Frozen, Result, Tags, Mask, and Hierarchy.  I’ve never seen “None” listed as a column type in any of my projects, but the Help menu says it’s an option.  It is important to note that if you create a calculated column with a transformation, the type reflects “Imported”, not “Calculated”.
  • Data Type – Self Explanatory
  • External Name — Users can and do change column names in the General tab.  The ExternalName shows the column name from the source.  I’ve found this to be very handy from time to time when I lose track of changes or am working on someone else’s project.
  • IsValid — Each column has a True or False value, but is the value is only relevant for calculated columns.  If the value is True, the column expression is valid.  If the value is False, the column expression is not valid.  False is bad.  That means the calculation is not working.  Mostly the user deleted a column used in the calculation.  Excel simply returns a #NA, but Spotfire returns a NULL, and the calculation becomes invalid.  These types of errors are hard to spot unless you know to review IsValid.
  • Origin — This metadata specifies the data table the column originates from, which is very important because this information is not provided elsewhere.  If you review Data Table Properties, Source Information, the dialog will list columns ignored, not columns added.  This isn’t terribly helpful.  However, you can easily insert column by reviewing the Origin column.

Next, let’s look at the General tab.

Column Properties Metadata


The general tab is used for two things.

  1. Changing column names
  2. Updating calculated, binned, or hierarchy columns

You may also Freeze columns here, but I have yet to encounter a scenario where I want to freeze a column.  Freezing columns locks the column and disallows editing. The Help menu says this function was created for scenarios where “you want to save the result from a calculation to prevent it from being overwritten or for performance reasons”.  However, freezing columns also embeds the table, which I rarely want.

Column Properties General



The Formatting tab is self-explanatory, but I want to point out a few things.

  • Formatting columns of data are faster than formatting each and every visualization.
  • Spotfire has a short number format that takes up very little space on visualizations.  It’s handy, and I use it a lot.
  • Users may select many columns of data (of the same data type) and format them at the same time.
  • Users can apply formatting from one column to other columns so the same steps aren’t repeated unnecessarily.

Column Properties Formatting


The properties dialog lists all Column Properties in a DXP and then notes whether a particular column is part of that property.  Go to the next screenshot for an example.

Column Properties Properties1

In this example, I have created a drop down property control called MyDropDown.  It has 4 columns in it.  Note, the drop down is a Document Property.  By default, Spotfire will place all columns from a data table in the drop down unless you create a Column Property to limit the columns.  In this case, I created a Column Property called MyDropDownOptions restricting the drop down to only four columns.  Review the previous screenshot to see MyDropDownOptions listed in the Properties tab.  The value is set to True for c.Total BOE.  It would be False for Well Name, Prod Date, etc.
Column Properties Properties2

Column properties can be added through this menu or also through the Information Designer.  When using the Information Designer, the column property is added directly to the Column Element (in the case of information links).


The geocoding dialog allows the user to specify that a column contains geographic information.  That information may be used for positioning data on a map.  This post is already quite long, so I won’t to get into geocoding details.

Column Properties Geocoding

Sort Order

When placing a categorical column of data on a visualization, such as a bar chart, the data sorts itself in the “Standard sort order”.  However, users often want to change this order, and that is possible here in Column Properties using the “Custom sort order”.

Column Properties Sort Order


Final Notes

Before I wrap up Column Properties, I also want to note that you can create Column Properties for individual column elements in the Information Designer.  That is a broad subject, and I’ve only done a few of these, so I’m not going to dive into it here.  However, this blog post on creating information links for spatial data provides two examples of adding column properties to the column element.  I would really like to learn what else is possible in this space.

And that wraps up Column Properties!  Next, I’ll discuss Data Function Properties.

Guest Spotfire blogger residing in Whitefish, MT.  Working for SM Energy’s Advanced Analytics and Emerging Technology team!

Unscrambling the Spotfire Text Area

  • Are you frustrated with the Spotfire Text Area?
  • Are you just getting started with HTML, CSS, JavaScript, or jQuery?
  • Do you find all the different languages that can be implemented in Text Areas confusing?

Read More

Guest Spotfire blogger residing in Whitefish, MT.  Working for SM Energy’s Advanced Analytics and Emerging Technology team!

Part 2 – Spotfire Properties – Data Table Properties

This is the second part of a series on Spotfire Properties.  In the previous post, I talked about Document Properties.  This week we’ll take a look at Data Table Properties.

Data Table Properties

As the name indicates, Data Table Properties controls everything related to data tables.  It’s organized into 6 tabs. No matter which tab you click on, the data table names and the four buttons on the right-hand side of the screen will always be visible.  You may notice each of the data tables has a square of color next to the name.  This is a visual indicator of whether or not tables are related.  If the tables are related, they will be the same color.  If the tables are not related, they will have different colors.

  • Rename — Allows the user to rename the table
  • Refresh Data — Click the button and then choose Refresh Data with or without prompt.  This button was CRITICAL before TIBCO added the ability to edit transformations from the Data Panel.  In the past, users had to use Refresh Data with Prompt to step through transformations, although this is no longer the case.
  • Delete — Allows you to delete tables.
  • Set as Default — When new visualizations are added, they are connected to the default table.  Change this to the table you use most frequently.

Data Table Properties General

The General tab has most of what users are looking for.

  • Store data — This radio button allows you to toggle between embedded and linked to the source.  Spotfire defaults to linked.  As long as it is linked, the DXP will ping the data source each time the file opens.
  • Key columns for linked data — This setting applies to tags and replacing specific values.  If you are new to Spotfire, check out this post on tags, which references this dialog.  Replacing specific values was a new function added in version 7.9.  Users may now double-click on a value in a table and replace that individual value if keys are set up in Data Table Properties.
  • Type of data — This box will appear gray and unavailable unless you are importing data On Demand.  If loading On Demand, this is where you can change On Demand settings.
  • Filters — By default, Spotfire will create a filter for each column of data in the filter panel.  You can change this setting by setting the radio button to ‘Manage manually’.  I talk about manually managing filters in this post.  Spotfire also defaults to caching columns automatically.  This setting can be changed in Data Table Properties or in Administrative Settings.  You can check out this post for more information on caching calculated columns.

Data Table Properties Source Information

Next, the Source Information tab displays the analysis build steps.  Before the Data Panel Source View, this was critical.  You may still need Source Information for more detailed information outside the Source View.  If I need to rebuild an analysis, I copy and paste this into Microsoft Word for reference.

Data Table Properties Source Info

Data Table Properties Relations

Relations facilitate marking and filtering across different data tables in Spotfire. If this case, I have two tables — Well Location Data and Well Producing Depth Data.  I have a common column between them that I want to filter on, so I create a relation to do that.  Click the Manage Relations button to see the relations.  This screen will be blank unless the user has created relations.

Data Table Properties Relations

Data Table Properties Column Matches

Column Matches are what allow users to place data from different tables on the same visualization.  Spotfire creates them for you when you have the same column (name and data type) in more than one data table.  For example, if I have a column called Well Name where the data type is a string in both tables, Spotfire creates the match.  In this particular case, my two tables didn’t have columns that matched the name and data type, but I created a column match manually on ComplName and Well_ID.

Data Table Properties Column Matchs


Data Table Properties Properties

Data Table properties are usually created in Information links and appear in the Properties tab.  Users may also create new properties here.

Data Table Properties Properties


Data Table Properties Scheduled Updates

Lastly, Scheduled Updates primarily impact analysis files used by the Web Player, as described in this blog post.  However, they can also assist in optimizing desktop files as well, as described in this blog post.  In this dialog, TIBCO provides a link to your server where you can set up scheduled updates for the DXP.  It also lists all source data tables contributing to the data table you have already selected above.  If you click one of the check-boxes, Spotfire excludes the table from the scheduled update and will reload when a new user opens the analysis.


Now, you know everything there is to know about Data Table Properties!


Guest Spotfire blogger residing in Whitefish, MT.  Working for SM Energy’s Advanced Analytics and Emerging Technology team!