JMP® is statistical analysis software with capabilities that span from data access to advanced statistical techniques, with click of a button sharing. The software is interactive and visual, and statistically deep enough to allow users to see and explore data.
$1,320
per year per user
MATLAB
Score 9.0 out of 10
N/A
MatLab is a predictive analytics and computing platform based on a proprietary programming language. MatLab is used across industry and academia.
Many organizations have seen their analytical capabilities, and the results from them, plateau. Of these, we've observed, that most of them didn't appreciate that they could do (even) better. These companies should definitely consider JMP. Any company that is research-based can benefit from accelerating their research, learning more in less time, effort and cost, with JMP's tools. Basically, any organization that is hungry enough for improvement to seek out better ways is suitable for JMP. Those who are happy with their current performance are not likely to consider the changes, though they were not major impediments by our clients, required.
Engineering, mathematical modeling, and machine learning are all fields where MATLAB will shine. It's fast, reliable, and relatively easy to use. MATLAB is the de facto standard when it comes to producing high-quality plots. If you need to deal with large data sets, and not take forever processing them, MATLAB may very well be the tool for you!
Has robust and easy-to-use debugging tools that can help one identify problems in one's codes.
Rich, well-developed and efficient library of mathematical and statistical functions that one might need to develop models or perform statistical analysis.
A very active online user community that is a great resource in terms of seeking help when you hit a snag.
Great help literature (and sometimes videos too) on all tools making it possible for all to train themselves.
MATLAB should have a full free version (without time limit) in order to be more accessible and thus have a greater user community.
The idea of having toolboxes to work directly with hardware (microcontrollers, single-board computers) is great, but one can tell it isn't updated very frequently and there isn't as much documentation available as with more common resources.
Our organization had a lot of trouble getting our network licenses to work properly and there wasn't any local service provider that could help us get it to work faster.
I've mentioned this earlier, but the licensing agreements are very prohibitive. I work with a company where my role has become less and less doing my own analytics and more and more trying to help other people in that role. As we are bringing more people "up to speed" it's hard to justify licenses for 2-3 people when they aren't full time, Six Sigma black belts just looking at stats all day. A floating license option would make this a no-brainer, since these people could continue their other work and add JMP usage as they grow their skills, but this is not something JMP/SAS has offered.
The GUI interface makes it easier to generate plots and find statistics without having to write code. The JSL scripting is a bit of a steep learning curve but does give you more ability to customize your analysis. Overall, I would recommend JMP as a good product for overall usability.
The best thing about MATLAB is the variety of research and development fields it supports. The reason for this rating is that it is best used for medical images enhancement and signal processing, it is also used for speech to text conversion. This tool server is best when the demand is for machine learning.
The helpful tips are great for new users. I am always able to find solutions to a tool I am working with through the hep section. And my area has a users group that meets each quarter to share ideas and view upcoming JMP revisions.
The built-in search engine is not as performing as I wish it would be. However, the YouTube channel has a vast library of informative video that can help understanding the software. Also, many other software have a nice bridge into MATLAB, which makes it very versatile. Overall, the support for MATLAB is good.
We actually use both JMP and IBM SPSS, but I think JMP's complexity lends itself to more in-depth statistical analyses. SPSS is designed for that as well, but we tend to use it more for quicker analyses, and we have found that JMP is far more powerful.
The commands and coding language of MATLAB reads a lot more in plain English as opposed to all the periods and other special characters that are needed when typing in Python or Java. Additionally MATLAB has several different function packages that can solve all different categories of problems so you don't have to make a bunch of different code scrips from scratch.
JMP has resulted in literally millions of dollars in ROI due to identification of correctable errors.
Use of JMP control charts JMP has greatly simplified and improved interpretation of Lean, FMEA, and PDSA type analyses.
Use of JMP has enable the testing and subsequent selection of 'best practices' saving uncounted hours in false starts based on 'collective experience'.
The down side is that JMP is not a 'magic box', one still has to take care in applying the tools properly. Moreover, time-consuming approaches using JMP may still be the 'order of the day', because the service (even power user) is unaware of significant shortcuts available for free on the JMP community website.
