Permissions continue to present mysteries and surprises. I recently came across this interesting situation: in the process of configuring the Project Permissions for a particular group it up and disappeared from the Permissions page.
The Tableau Public workbook below shows what happened. Follow the dashboards #1-#4.
Christian Chabot, in his TC 2014 keynote, likened Tableau to digital photography vs traditional business data analysis' film.
It's an excellent metaphor, capturing the essence of how Tableau transforms the nature of the relationship between people and the data that matters to them.
I'm glad I was able to help.
The original post, March 19, 2013
Linked in "Best of the Tableau Web...", March 27, 2013
Echoed by Tom Brown, The Information Lab,Oct 28, 2013
It was an inspirational day at TC 2014. Better than I expected, or deserved.
I sat in on Jonathan Drummey's "Extreme Data Blending" session, and was amazed at the depths he's plumbed in ferreting out the mysteries of how data blending works. Even better was the clarity of his presentation, making the complex and esoteric seem familiar and graspable. I'm now much better equipped to employ data blending to good effect than I was a day ago.
Sadly, I missed Joe Mako's "Flowing With Tableau" session. Although I got there 20 minutes or more before it was to start the line was already far longer than would get into the room, and many people were left bereft. This was a disappointment, as Joe's and Jonathan's sessions were the two I most wanted to attend. I remember the first Joe Mako session I went to, some years ago, to see this fellow I felt I knew from his Tableau forum support. After the session I went up an introduced myself, feeling like a little boy meeting Wayne Gretzky (I'm Canadian), and to my surprise Joe was a friendly in person as he was online.
I also had the great pleasure of meeting Richard Leeke, whose mapping work I've been borrowing for a long time, and Noah Salvaterra, who's blazed across the Tableau firmament creating astonishingly beautiful and illustrative works with Tableau that almost seem impossible, or at least dark magic.
At one point I was in the company of these four, and Zen Master Ramon Martinez, whom I had the pleasure of working with at the Pan American Health Organization some time ago. I felt a little bit of an interloper, hugely lucky, and something of a fanboy in the company of the smartest, most clever, and most generous members of the Tableau community.
It's easy to get lulled into the day to day routine. When your work is steady state, and constant, there's the tendency to become comfortable with the familiar, and the allure of the interesting puzzles to gnaw on dims. After this conference, after seeing the remarkable things I seen people do, people whom I have the utmost respect for, I'm energized again and looking forward to pursuing new and interesting avenues. There's a lot of Tableau left to cover, and there's no time to waste.
or, Is my 'Profit' the same as your 'Profit'?
It's helpful to be able to identify what fields are being calculated in Tableau. From an enterprise governance perspective it's essential.
Unfortunately Tableau doesn't provide this information except through manual inspection of the fields' calculations in the calculation dialog. This is a serious deficiency, one that hinders Tableau's adoption in the Enterprise, where senior management is reluctant to let Tableau loose "in the wild" for fear of having dueling metrics without transparency.
Fortunately, there's an easy solution. This post presents a simple Ruby script I whipped up that examines Tableau workbooks and creates a CSV file containing the relevant information about the calculated fields. This CSV file is then usable with Tableau to provide the transparency necessary to tell whether or not two metrics are in fact the same.
Example: the Tableau Sample Workbooks' Calculated Fields
As an illustration of somewhat real world calculated fields, here are those found in the Tableau v8.2 sample workbooks. They range from simple to reasonably complex.
Here's how it works in the simplest case: you copy/paste the code below into file named oh, say "TTC_CalculatedFields.rb" — it's then a bona fide Ruby script. Then from a command prompt run it in from the directory containing the TBWs you want the calculated fields from.
The usual way to run it is with
"{pathtoRuby}\Ruby" TTC_CalculatedFields.rb
It will then find all of the TWBs in the directory, and all the sub-directories, and record all of their calculated fields.
The usual caveats apply.
On the other hand, it should be pretty easy to address any shortcomings it has. Ruby is an elegant language that makes fixing and improving things generally simple and straightforward.
# TTC_CalculatedFields.rb - this Ruby script Copyright 2013, 2014 Christopher Gerrard
require 'nokogiri'
require 'open-uri'
$recNum = 0
$CSVHeader = <<CSVHEADER.gsub("\n",'')
Record Number,
Workbook,
Workbook Dir,
Data Source Name,
Data Source Caption,
Data Source Name (tech),
Name,
Caption,
Name (tech),
Data Type,
Role,
Type,
Class,
Scope Isolation,
Formula Code,
Formula,
Formula Comments
CSVHEADER
def init
$f = File.open("TTC_CalculatedFields.csv",'w')
$f.puts $CSVHeader unless $f.nil?
end
def parseCalculatedFields twbWithDir
twb = File.basename(twbWithDir)
twbDir = File.dirname(File.expand_path(twbWithDir))
puts "Found '#{twb}' in #{twbDir}"
doc = Nokogiri::XML(open(twb))
calculationNodes = doc.xpath("//workbook/datasources/datasource/column/calculation").to_a
calculationNodes.each do |d|
#-- Data Source --
dsTechName = d.xpath('../../@name').text
dsCaption = d.xpath('../../@caption').text
dsName = if dsCaption == '' then dsTechName else dsCaption end
#-- Field --
caption, = d.xpath('../@caption').text
techName = d.xpath('../@name').text
name = if caption == '' then getName(techName) else caption end
dataType = d.xpath('../@datatype').text
role = d.xpath('../@role').text
type = d.xpath('../@type').text
#-- Formula --
formulaCode = d.xpath('./@formula').text.gsub(/\r\n/, ' ')
formulaLines = d.xpath('./@formula').text.split(/\r\n/)
formula = getFormula( formulaLines )
comments = getComments( formulaLines )
fClass = d.xpath('./@class').text
scopeIsolation = d.xpath('./@scope-isolation').text
#-- emit CSV
putCSV(twb,
twbDir,
dsName,
dsCaption,
dsTechName,
name,
caption,
techName,
dataType,
role,
type,
fClass,
scopeIsolation,
formulaCode,
formula,
comments
)
end
end
def getName techName
techName.gsub(/^\[/,'').gsub(/\]$/,'')
end
def getFormula lines
formula = ''
lines.each do |line|
line.strip
formula += ' ' + line unless line =~ /[ ]*\/\//
end
return formula.strip
end
def getComments lines
comments = ''
lines.each do |line|
line.strip
comments += ' ' + line.strip if line.start_with?('//')
end
return comments.strip
end
def putCSV(twb,
twbDir,
dsName,
dsCaption,
dsTechName,
name,
caption,
techName,
dataType,
role,
type,
fClass,
scopeIsolation,
formulaCode,
formula,
comments
)
$recNum += 1
csvLine = <<CSV.gsub("\n",'')
"#{$recNum}",
"#{twb.gsub('"','""')}",
"#{twbDir.gsub('"','""')}",
"#{dsName.gsub('"','""')}",
"#{dsCaption.gsub('"','""')}",
"#{dsTechName.gsub('"','""')}",
"#{name.gsub('"','""')}",
"#{caption.gsub('"','""')}",
"#{techName.gsub('"','""')}",
"#{dataType.gsub('"','""')}",
"#{role.gsub('"','""')}",
"#{type.gsub('"','""')}",
"#{fClass.gsub('"','""')}",
"#{scopeIsolation.gsub('"','""')}",
"#{formulaCode.gsub('"','""')}",
"#{formula.gsub('"','""')}",
"#{comments.gsub('"','""')}"
CSV
$f.puts csvLine unless $f.nil?
end
init
Dir.glob("**/*.twb") {|twb| parseCalculatedFields twb }
$f.close unless $f.nil?
An interesting exercise would be to round up the various workbooks that have been published about Table Calculations to see what that population looks like. Maybe something to do during the conference this week.
I hope you find this little utility useful. If so, please leave a comment. If it doesn't work, please leave a comment — I may or may not be able to help (time, other demands upon, etc.). If you extend it in some useful way, please leave a comment.
Or just leave a comment.