slam package updated --> PubMedWordcloud cannot be installed without it

[cnobles]

The package was updated to version 0.1-38 on Aug 18th 2016. This is a dependency for the packages 'tm' and 'wordcloud', which are in turn dependancies for PubMedWordcloud.

We can enforce package version by installing from the CRAN archive rather than default.

Lines below can be inserted into the setupINSPIIRED.R script.

slam_url <- "http://cran.r-project.org/src/contrib/Archive/slam/slam_0.1-37.tar.gz"
install.packages(slam_url, repos = NULL, type = "source")

[chasberry]

Why hard code the version? I'd do this:

slam_url <- "http://cran.r-project.org"
install.packages("slam", repos = slam_url, type = "source")

[chasberry]

Or to say this another way, I'd address the dependency issues rather than work around them.

[cnobles]

If we started using the latest version of R (3.3.1), many of the other dependent packages would not work just yet as they have not been updated. For INSPIIRED, we've elected to use R 3.2.2 for which every packaged is compatible. It seems that the only way to be certain the program will be able to run as we currently have it designed would be through version control. In the long run, we may need to version control all the dependancies, and update them only upon another release of INSPIIRED (v2.0).

Do you have any recommendations or is this a good course of action?

[chasberry]

Not following this:

many of the other dependent packages would not work just yet as they have not been updated

I see them on CRAN where it appears that they pass R CMD check. So they look up to date to me and I'd just use them.

If you truly cannot update, then I'd ask see if R-devel has a best-practices given the particulars. And then if there is nothing, I'd post a query.

[cnobles]

I tested transitioning to R 3.3.1 yesterday with our installation management software (conda), but ran into some issues that I'll try to resolve. I'll look into making that transition more and see if version control is really necessary. Thanks for the suggestion.

[navpreet2130]

i am using R 3.2.2 i tried to iinstall slam package with code
slam_url <- "http://cran.r-project.org"
install.packages("slam", repos = slam_url, type = "source")
bt it gives error: installation of package ‘C:/Users/User/AppData/Local/Temp/RtmpyyeUNs/downloaded_packages/slam_0.1-37.tar.gz’ had non-zero exit status
please help

[cnobles]

Try pasting the two lines of code at the beginning of this post into line 20 of INSPIIRED/bin/setupINSPIIRED.R before running the INSPIIRED setup. That should install the package first thing.

[navpreet2130]

Hello sir. I have installed all the packages . I am doing sentiment
analysis on Twitter data and want as output the following graph as shown in
the picture attached. Please help me to do this .
On 29-Sep-2016 11:17 pm, "Chris Nobles" notifications@github.com wrote:

Try pasting the two lines of code at the beginning of this post into line
20 of INSPIIRED/bin/setupINSPIIRED.R before running the INSPIIRED setup.
That should install the package first thing.

—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
#2 (comment),
or mute the thread
https://github.com/notifications/unsubscribe-auth/AVSRRZKXJzy3AtOAncIDcOelwoGuAVjbks5qu_nKgaJpZM4Jp20l
.

[helixscript]

I would recommend installing slam through the conda environment.

From inside conda environment:
conda install r-slam

On Thu, Sep 29, 2016 at 1:51 PM, navpreet2130 notifications@github.com
wrote:

Hello sir. I have installed all the packages . I am doing sentiment
analysis on Twitter data and want as output the following graph as shown in
the picture attached. Please help me to do this .
On 29-Sep-2016 11:17 pm, "Chris Nobles" notifications@github.com wrote:

Try pasting the two lines of code at the beginning of this post into line
20 of INSPIIRED/bin/setupINSPIIRED.R before running the INSPIIRED setup.
That should install the package first thing.

—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
<#2 (comment)
,
or mute the thread
<https://github.com/notifications/unsubscribe-auth/
AVSRRZKXJzy3AtOAncIDcOelwoGuAVjbks5qu_nKgaJpZM4Jp20l>
.

—
You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub
#2 (comment),
or mute the thread
https://github.com/notifications/unsubscribe-auth/AHSJQ6FirVTxO6bN9J4dvO4ILy6ZbM6Iks5qu_q-gaJpZM4Jp20l
.

John K. Everett
http://www.linkedin.com/in/JohnEverettPhD

[cnobles]

Likewise, that should work too. I believe it was available the last time we checked. Sry, no picture was attached.