Portable R for Bloomberg
BloomR is a portable R distribution reshaped in order to get, manipulate and run models against Bloomberg data.
Focus on Data
Connecting R and Bloomberg is difficult and time consuming. BloomR has no learning curve: extract its files on your USB flash disk, plug it into any Bloomberg terminal where you have access and you are done.
If you prefer to copy BloomR files on your Bloomberg terminal, that's fine too.
Work with Data
BloomR has a number of functions to easily get and manipulate Bloomberg data: for example, with a single line of code you can get the historical data from all members of a given index in R xts format.
To save you time, BloomR functions can work in bulk with many tickers together, possibly grouped in categories that you define, representing your portfolios, sectors, etc.
Visualize and Communicate your Research
BloomR can save the output of your models in spreadsheet format and can produce template based reports in HTML or PDF format. You may also want to export model plots as figures for your presentation.
In a business environment where you have to produce a lot of similar template based model, for compliance or reporting purposes, BloomR can feed you templates with up to the minute fresh data and release it as a HTML or PDF: the former can be delivered with the company web site, the latter can be circulated as a document.
Integrate with Excel
Your research might need integrations from other data other colleagues or departments. Chance there are they will feed you with Excel data. BloomR can read Excel xlsx files. It can import data from specific Excel sheets or ranges.
In particular, you might want to minimise the risk of errors by using Excel named ranges, e.g. BloomR will just import the Financial_Ratios named range from you peers and you don't have to guess where needed items might possibly be located inside the workbook.
Decoupling the Model Development and Production
You can run BloomR alone (without a Bloomberg terminal) in simulated mode. In this way you might design your models and tests with artificial data and, when you think you ready for production, you plug it into a Bloomberg terminal and test your models with actual data.
Reproducible Research
BloomR was born primarily as a research tool. The process of manually download lots of ticker data through a spreadsheet and manually accommodating into your sheets, besides being time consuming, is known to be prone to subtle undetectable errors. In BloomR this can be accomplished with a script so that you, your team, your colleagues can exactly reproduce it, fix it, improve it.
In a business environment this also facilitates the process of keeping track and documenting the data mining processes.
No Extra Licensing Restrictions
BloomR itself is open source and it is based on the public API released by Bloomberg (Bloomberg Open API). This means that the same licensing terms of your Bloomberg Professional Service apply to data when you work with it.
Improvements over Début version:
- BloomR development is now open source on Github.com.
- Only 64 bit (it means less space).
- Updated to R-3.2.0 .
- Green installer, to simplify extraction of BloomR files.
- Generation of automatic HTML reports for web publishing or in PDF (PDF's require a separate large add-on to be downloaded separately)
- New functions for bulk download using tickers from CSV files, from index constituents or from ticker vectors.
- Simulated download allows to prepare scripts at home and test on Bloomberg terminal later.
- Bulk and non-bulk download of many description fields.
- More consistent naming scheme:
bbg.*suffix is deprecated now, usebr.* - A formal documentation (see bloomr.html or bloomr.pdf)
- Demo and examples.
- BloomR can be downloaded as a binary or generated from an R script. So you can stay always updated with the latest version of R (and related packages).
read.xlx
read.xlx joins the family. With this function you can read Excel xlsx workbook sheets into R data frames. Some features are:
- It can import all, one, or a selection of sheets, where specific sheets are requested by means of their name.
- Instead of importing all the sheets' cells, it can import only those comprised in a named range.
- It can distinguish between cells formatted as numbers, percent, text and dates,
- Date cells are recognised whatever the language locale.
- Blank (visual) lines are detected and automatically removed from the data frame, unless you want to keep them.
- The filter is not based on any external engine and does not requires Excel to be installed at all. It's pure R code, so you can read xlsx files on Linux systems.
See xlx.help.html, or xlx.help.pdf user manual for more.
The new version of the BloomR manual for lecturers (in slide format) is coming soon.
Find in your BloomR directory the subdirectory help intended for learning and reference material.
For a self-paced introduction, download BloomR Classroom Companion from the latest BloomR release.
-
Download the latest BloomR release.
-
Run the green installer: it will only extract the BloomR files on your chosen directory (perhaps on your USB drive), without further touching your system.
-
Run "BloomR.exe" in the main folder.
-
Log to Bloomberg service on the same PC (if you are not logged already).
-
To get some data, type in the BloomR console:
con=br.open() # Connect BloomR to Bloomberg br.bulk.tiks(con, c("MSFT US", "AMZN US"), addtype=TRUE) # Last 5 days for price of these tickers br.bulk.idx(con, "SX5E Index", field="PX_LAST") # Last 5 days for price of these indices br.close(con) # Disconnect BloomR from Bloomberg
This section is intended only for developers willing to build BloomR themselves. Standard users will not need this.
Download bloomr.build.R (as a text file, not as a HTML file, and make sure your browser does not change ".R" extension to ".html" or ".txt"). In the console of R for Windows issue:
source("path\to\bloomr.build.R")
makeBloomR("path\to\workDir")
You will find your BloomR green installer .exe inside workDir.
Note: the build script will download from the Internet the required files (c. 200 MB). Think of this if you have a metered connection.
Browser oddities
Downloading a file as is from you browser can be tangled. Normally you need to a) select All Files in the Save-as option of the browser Save dialog and b) double-quote the filename. In as far as the format is text (not HTML), even if you get a distorted name, like "bloomr_build_R.txt", you can always adjust the source()-path accordingly.
Yes, browsers are strange sometimes.
A decent R version is needed 3+ (perhaps older version will do).
.Platform$OS.type == "windows". XML and Rcurl packages are needed, but if missing, the build script will try to download and install them.
R should be able to connect to the Internet.
If you want to find BloomR.zip too in your work directory use:
makeBloomR("path\to\workDir", zip = TRUE)
Since bloomr.build.R downloads resources from the net, one problem you might face is a download error.
Based on my experiences, the script does a good job in detecting download errors, despite it is always possible that a corrupted download is not identified.
When a donwnload error is detected, bloomr.build.R makes a second attempt. You can control this with:
makeBloomR("path\to\workDir", ndown = n)
where n are the download attempts.
When the maximum number of attempt is reached, the script exits deleting any incompleted file and emitting a message detailing the download error.
When you start over (hoping in a better connection), you might not want to download again every and each file, but only those affected by the errors. This is done with the option:
makeBloomR("path\to\workDir", tight = TRUE)
When you run BloomR installer in Windows 10, you might might presented with the following warning:
In order to remove it, I should buy a certificate, which by the way would not make the BloomR installer safer, it would just make simpler to identify me, Antonio Fasano, as the author of the project.
The certificate costs money, while this project is not for profit and does not generate any.
Anyway, BloomR is open source, so you can inspect the code to see if there is any wrong doing and, of course, you can run an antivirus program on the BloomR installer before running it.
Bear in mind that the installer does not require administrator privileges to be executed because it does not affect critical system services.
All that given, to execute the BloomR installer, identify the More info link as displayed in the following screenshot:
After clicking on the link, you get the following:
If you want to run the executable, click Run anyway.