topic Re: worth to recommend the use of R statistics in ArcGIS in R-ArcGIS Questions

worth to recommend the use of R statistics in ArcGIS

larryzhang — Sat, 30 Jan 2016 18:11:37 GMT

With 10.3.1 /Pro 1.1, we can eventually integrate open-source R statistics with ArcGIS (Python scripting) for wide data analytics, in particular, combined with R-ArcGIS Bridge (R-ArcGIS/r-bridge-install · GitHub ).

If high-performance RRE (Revolution R Enterprise) is used, 64-bit ArcGIS /Pro can easily be extended to make use of the R distributed computing infrastructure.

If any experience with this aspect of practice, please share

Re: worth to recommend the use of R statistics in ArcGIS

DanPatterson_Retired — Sat, 30 Jan 2016 18:49:46 GMT

How does it differ from the announcement here....

Building a Bridge to the R Community | Esri Insider

and the repository here?

R — ArcGIS

A comparison would be useful as to what your link offers.

Re: worth to recommend the use of R statistics in ArcGIS

ChadKopplin — Sun, 31 Jan 2016 03:16:53 GMT

Please keep us informed it would be great to hear what you find if you do a comparison, like Dan said.

Re: worth to recommend the use of R statistics in ArcGIS

larryzhang — Sun, 31 Jan 2016 09:21:52 GMT

After quick browsing it, it looks same as what I recommend here .

Good move, ESRI!

Personally, R Bridge with ArcGIS is a good start for data engineers (with Python coding skill) to 'directly' call rich open-source R statistics packages inside ArcGIS.

In fact, R (and SAS) is an advanced (standard, powerful) statistics tool for data engineers for over 20 years, in particular, in governments and larger business organizations, because of rich algorithms and great visualization capabilities offering huge potentials on data analytics, including spatial data (vector and raster).

Even though statistics algorithms / libraries in Python have been gaining attractions (incl. ESRI ArcGIS), majority of data engineers still prefer R (SAS or both) against Python, in practice, because of many and many reasons (you can google-it) ...

Re: worth to recommend the use of R statistics in ArcGIS

DanPatterson_Retired — Sun, 31 Jan 2016 09:26:23 GMT

Ahhh the disciplines... R is a good inclusion and there is good overlap and differences with SciPy amongst other. Pretty soon, everything will be included... not to forget other programming languages and libraries like The Julia Language

Re: worth to recommend the use of R statistics in ArcGIS

DanPatterson_Retired — Sun, 31 Jan 2016 11:32:17 GMT

Modules aside... there is an interesting comparison of some of the capabilities from a programming perspective between, the most common 3 languages (Matlab, R, Numpy (python/scipy/pandas) and the relative noob ... Julia

Numerical Analysis & Statistics: MATLAB, R, NumPy, Julia - Hyperpolyglot

Re: worth to recommend the use of R statistics in ArcGIS

larryzhang — Sun, 31 Jan 2016 13:34:01 GMT

As a reference, in practice, four main languages (R, SAS, Python, SQL) for data analytics, data mining, and data science, were "generally" surveyed in 2014 (even though some are using EXCEL, MatLab or SPSS), where data include spatial data (vector, raster), in addition to numerical and text.

Re: worth to recommend the use of R statistics in ArcGIS

DanPatterson_Retired — Sun, 31 Jan 2016 14:37:05 GMT

And of course this rules out C, C++ and lest we forget... Fortran

Re: worth to recommend the use of R statistics in ArcGIS

SteveKopp — Fri, 24 Jun 2016 16:03:18 GMT

For contributors and followers of this thread - there is now a GeoNet group "R-ArcGIS" dedicated to people working with both R and ArcGIS. This will provide a centralized place for sharing of ideas and instructional materials. We will continue to use the GitHub site for distribution of the R-ArcGIS bridge installation, and access point for all related source code.