We often find ourselves helping people start a DPL analysis from an Excel spreadsheet, but in many cases the difficult modeling is in the spreadsheet itself. In this post I'm going to provide you with a few simple tips for preparing your spreadsheet in a way that makes setting up and running a scenario analysis a breeze.
Do you find it annoying that your model consists of both a DPL workspace file (e.g., FOO.DA) and an Excel workbook (FOO.XLSX)? Well, don't despair, embed!
The DPL maintenance update 8.00.10 is now available for download. It includes the following:
CHG: Better install errors in DPL Direct (#1879)
BUG: Updating a tornado can crash if certain objects are selected (#1887)
BUG: Crash in initial decision alternatives productivity ratio chart with zero differences BUG: Objective function label not display on risk profile chart (#1852)
ENH: Allow an event to appear more than once on a tree path and halt (#1851)
The humble tornado diagram, long a staple of risk and decision analytic methods, is disarmingly simple, but it can be surprisingly difficult to do right. Tornadoes are on my mind lately because we've recently published a video on them:
May's Model of the Month was submitted by a group of ParisTech's GRID students who, under the guidance of long-time DPL user Professor Marc Lassagne (http://www.msmgr.fr/bio-des-responsables/), developed a model that will help practitioners choose the most appropriate type of hip prosthesis based on patient demographics.
The model uses 3 decisions (Material Type, Fixation Type, and Prosthesis Type) to enumerate the best overall prosthesis choice for a particular subset of patients based on the following criteria:
How do you tackle a model with 318 paths? Well, whatever you do, you better not do it 318 times! Even if you were able to calculate a billion paths a second, it would still take 30,000 years or so.
The model we recently shared as part of the March Madness promotion was solved using Hybrid Discrete Tree Simulation, one of DPL's decision tree evaluation methods.
Ask yourself: When you strive for perfection in decision making, when you go the extra mile to hunt up that critical assessment or wrack your brain in search of yet another creative alternative, is it possible you're just trying too hard?
If you have school-age children, you are probably well aware that Saturday is international Pi day. Yes, 3/14/15 bears a passing resemblance to 3.14159265... Now, the skeptics among you will note how weak this resemblance is:
The DPL maintenance update 8.00.09 is now available for download.
BUG: Node definition dialog not modal when creating a chance node for a single row in DPL Direct (#1848)
BUG: Compile may fail after extracting database data containing strings with backslashes (#1847)
BUG: Fixed "event already active" error with VOI on chance nodes dont gambled in some but not all decision tree node instances (#1835)
A recent post by Steve Barrager on LinkedIn got me thinking about decisions, and in particular, the framing choices that lead to the identification of the focus decisions an analytical project will consider.
In many cases, the main decision is pretty obvious. Should we fund this R&D project? What should we bid for this lease?
All versions of DPL 8 include DPL Direct, an easy-to-use yet robust decision modeling environment built as an add-in utilizing Excel's ribbon interface. The DPL Direct version offers the Excel add-in environment only, while the DPL Professional, Enterprise, and Portfolio versions offer users a choice between building DPL models within Direct (i.e., Excel) or the standalone environment.
The DPL maintenance update 8.00.08 is now available for download. This update includes several enhancements and bug fixes, including these:
ENH: New buttons on the DPL tab in the Excel side interface for examples list and quick start guide (#1814)
BUG: Attributes with similar prefixes may be erroneously renumbered in objective dialog (#1805)
BUG: Find may skip some strings with whole word option on (#1807)
ENH: Weibull distribution now available for continuous Monte Carlo simulation (#1809)
BUG: Display of endpoint database may result in outputs being cleared (#1769)
First, let me note that up here in Maine we are getting another storm that is supposed to dump a foot of snow on us. This one doesn't have enough wind to qualify as a blizzard, but snow is snow. Second, I'd like to add a few notes about the previous storm, the blizzard that socked most of New England but spared New York City.
If there are two things that get us excited here in Syncopation's Bangor, Maine office, they are probability and SNOW! Thus, we were all very pleased to see this graphic from the National Weather Service's Caribou office posted on twitter:
We don't usually use this space to comment on swings in the commodity markets, but the recent and ongoing plunge in oil prices is hard to ignore. For people in the energy industry, the news is generally bad, whether they're in petroleum or its alternatives. For most others, the news is good, akin to the economic tailwind provided by a broad based tax cut.
The US economy seems to have finally founds its footing, yet budgets for software training and associated travel costs remain at recession-level. At the same time, web based learning applications and technologies have taken off.
I have to admit I often moan and groan when companies quoted in the press make stupid statements about uncertainty, or reveal that their decision making is horribly risk averse to the detriment of their shareholders/stakeholders. Conversely, I feel almost obligated to note the occasion when a business leader speaks intelligently about risk. In this case that would be Elon Musk of SpaceX.
While we normally try to avoid dry technical topics in this blog, the time has come to say a few words about 32- and 64-bit applications, in the hopes of addressing any confusion you may have registered. In this post, I'll answer some key FAQ's and provide a few pointers to other sources of useful information.
Q. Do I really need to know about this?
