Origin and OriginPro 9.0


Origin and OriginPro (hereafter, just ‘Origin’ with indication where something is only available in the Pro version) both went 64-bit in their previous release (8.6) and it was a welcome advance, but this is the update in which the effects really start to show off. On the most intensively demanding graphing tasks you don’t need a benchmark test to show their speed gain; it’s clearly visible as you work. One complex 3D operation which used to allow me time for a swig of coffee is now complete before I lift my finger from the mouse click which triggered it.


During the period of review, I worked with researchers who were using release 8.6 as part of a workflow pipeline for text analysis. Replicating some parts of their technique at home it was clear that, even in this untypical mixed role, things ran faster in the new version, even on a slower machine. Exactly how much of this is a result of refined 64-bit utilisation and how much is down to default OpenGL graphics for 3D work I can’t say, but the net result is startling and impressive.


Putting aside the speed issue, my personal favourite picks were additions which make work flow more smoothly and, therefore, render data exploration and interpretation more transparent.


One of these, the latest addition to Origin’s repertoire of ‘gadgets’, is a comparative cursor. Origin calls it a global vertical cursor, since it runs comparisons across linked stacks of graphs panels selected from any disparate data sets the user wishes. Being able to flick inspection focus from one plot or group of plots to another, precisely comparing and contrasting x and y coordinates in a tabulated display, is not only a time saver but a valuable tool for expanding exploratory views. It jigs in perfectly with Origin’s earlier RoI (region of interest) inspection gadget a couple of releases ago.


Another of my favourites is the addition of an Excel-like data filter. Though there are slight differences in its look and feel, anyone who has used Excel’s column filters will feel right at home with Origin’s implementation. As a practical example, I had a dataset just now in which I had to analyse data only for those respondents who were aged over 18 and had answered ‘yes’ to a particular question. A couple of clicks hid the rows which were not required and, in an improvement over the usual spreadsheet approach, asked me whether or not I wanted to include the hidden rows in subsequent analyses. This could, of course, have been done in previous versions; it could also be done in other ways, but not with so little effort or fuss.


Making Excel users (almost all of us, to at least some extent) feel at home is a long-term aim of most worksheet-based software tools, and Origin is no exception. Also in 9.0 is an unremarked but significant adjustment in its approach to workspaces: a relaxing of the page-based paradigm. It comes in the form of floating plots which can be placed, as in many spreadsheets, within the worksheet itself rather than a separate child window.


There are many other welcome little touches scattered throughout, which seem minor but can make a disproportionate contribution to productivity and/or usability. The option to choose whether copy actions in the Data Info tool embrace the whole window (as previously) or just the chosen cell is one example (but make sure that you have the latest service release, as this was not available initially).


Moving to developments in the Origin’s central purpose, implicit 3D function fitting and infinite impulse response filtering (Butterworth, Chebyshev I & II, elliptic) are welcome additions to the Pro version. So too are a range of changes at the detail level, such as (again, as long as you have the latest service release) an option to sort segments by size within the columns of a stacked bar plot as an alternative to traditional parallel stripping. Direct plotting of bars and surfaces from raw (x, y, z) data points instead of the usual palaver with transformation to a grid also wins my love and affection. Negative values in radar plots and small value aggregation in Pareto, which I had overlooked, both won squeals of joy from colleagues who came in to play.


There were at least three individual features of this release that I would, even individually, see as self-funding justifications for upgrade. Taken together, for anyone depending on software of this type for their work, I’d say that the case is clear cut.



Sophia Ktori highlights the role of the laboratory software in the use of medical diagnostics


Gemma Church explores the use of modelling and simulation to predict weather and climate patterns


Robert Roe reviews the latest in accelerator technology and finds that GPUs and coprocessors will be key fixtures in the future of deep learning


Robert Roe finds that commoditisation of flash and SSD technology and the uptake of machine learning and AI applications are driving new paradigms in storage technology.