Entering edit mode
7.7 years ago
thomas.f.hahn2
▴
100
I am in urgent need to install R and R studio on a Linux server? How can I find out what Linux version the server has, what is the ftp download and install command for R and R studio and how do I start these 2 programs after I have finished installing them? Could you please respond fast because I am under extreme time pressure to get this done quickly? You can also email me instructions how to do this to Thomas.F.Hahn2@gmail.com Thanks a lot in advance for your help. Best, Thomas
If I a already on a huge server where I know R is somehow installed on it, what do I need to do to find out where it is and which file do I need to click on to get it started?
If I have the ftp download command, do I need to be in the directory into which I want to install R?
How can I need to connect R Studio with its latest R version under Linux? Please not that I only know two Linux commands, i.e. cd .. and ls. I use winscp when trying to find things in lLinux but it has not worked well.
What do you get what you type these two commands?
which Rwhich rstudioOn certain systems
whichisn't available, if this is the case for you OP, you can also usecommandor a load of others listed here: http://unix.stackexchange.com/questions/85249/why-not-use-which-what-to-use-thenYes. It seems to be R and R Studio together but we could not figure out the exact version because we could not get it started.
We've told you this several times now. To start
Rat the command line, you literally just typeRand then hit enter/return. Your shell will change and it will be obvious that R is running. Rstudio is more difficult because it's a GUI. But until you give us some actual information about the server config we cannot help you. You cannot see, but we cannot read minds either.I got R to run but the person, who came with me via TeamViewer to the server told me that on this kind of server R Studio cannot be install. R is kind of boring under Linux because it has no output console. That is why I cannot see in real time what my script is doing at any given time and stop it if it is not doing what I want it to do. Instead we get a huge output file, which takes forever to get read in the vi-editor.
Commandline R literally only has an output/input console so I dont really know what you mean. The only thing you can't do as easily on the command line as in Rstudio, is view plots and environment elements. I dont know who was helping you but I'm pretty sure you can install R studio on any server so long as xterm is correctly configured.
I can tell you on my laptop because that machine I understand.
Can such in installation be completed within one hour? Once I have R and R Studio running, how can I keep them both running and working hard ton completely time and computationally intensive scripts, which take more than a week on my laptop, on the server instead but without me needing to be connect to the server all the time?
Which program, .e.g. Putty, SSH or WinSCP would be best for this purpose and how do I need to configure them so that they will work?
Pleast STOP posting new questions as "answers". See my answer below.
It is very hard for me to navigate this communication interface properly because font color and background color are of low contrast. Therefore, I can only reply by randomly clicking the area until a new window opens, in which I can type something.
I just want the ready to use command to copy-paste and complete the task. I cannot read much since I am legally blind. These sites would take me weeks to read them all.
While I full sympathize with your predicament what you are asking for is above and beyond what volunteers at Biostars can do for you. We have no idea what server you are using/what you are able to see so in that sense we are in the same situation as you are.
Your best bet is to find a user on the server you are using and have them work with you.
I think I found everything to complete the task except that I cannot figure out the exact version of the Linux server.
Curious why you need that?
I think the information below pertains to my Linux version:
[tom@binf-dev ~]$ lsb_release -aLSB Version: :base-4.0-amd64:base-4.0-noarch:core-4.0-amd64:core-4.0-noarch:graphics-4.0-amd64:graphics-4.0-noarch:printing-4.0-amd64:printing-4.0-noarch Distributor ID: CentOS Description: CentOS release 6.8 (Final) Release: 6.8 Codename: Final [tom@binf-dev ~]$
But I can I now find the ftp download command that installs exactly the kind of R and R Studio for this particular Linux version. I watched somebody doing it. He was very good and it looked so easy since he knew what he was doing. He was training me but I forgot the command. I only remember that it could easily be donw.
Give us the output of these two commands first: C: How can I find out the Linux version and what is the ftp command to install R an
You may already have R and rstudio installed on this server.
I have access to two servers. One the big one R is already installed and I only need help figuring out how to get my scripts on their R. On the small server I can install program but I know R is definitely not on it because I manually went with WinSCP into each and every directory and could not find it. Unless it has a different name and I just did not recognize it. What is the command with which I could search an entire server to figure out whether R is already installed. Maybe you are right and its all there but since I have never seen it I don't know what I need to look for.
do what genomax said in the comment above. There's no point us helping you until we know what we're working with.
And why can't you just run your programs on the 'big' server?
I found on the website, which lists all the programs that are installed on Karst, which is a Linux Supercomputer for which it took me many hours to figure out how to log in this past night, that a relative recent version of R is installed on it. The link to this website, which talks about the supercomputers that I am working very hard on learning how to use, is https://cybergateway.uits.iu.edu/iugateway/moduleInfo?name=r.
There are three supercomputers, into which I can log into. Their names are Karst, Mason and Big Red II. Since I don't know how to query using regular expression on command line, I used WinSCP to look into as many directories as possible but the newest version of R, which I could find, was 2.18. But the Biocounductor Package, which I need to use seems to require at least R 3.3. And even though I could find R 2.18, it did not help me because I could not figure out a way to launch it.
The name of the Bioconductor Package, which I must use, is "TFBSTools". See http://bioconductor.org/packages/release/bioc/html/TFBSTools.html
I have been working all day yesterday running all the commands listed in the vignette for this package, which has the title: "Transcription factor binding site (TFBS) analysis with the “TFBSTools” package" and can be found at http://bioconductor.org/packages/release/bioc/vignettes/TFBSTools/inst/doc/TFBSTools.html, trying to understand what this package is actually doing. But despite having followed all the instructions I keep getting exactly the same error message when I run the line:
PFMatrixList <- getMatrixSet(JASPAR2014, opts),
which is the 7th lines of code under heading 3.1, which is titled "Search JASPAR2014 database". This JASPAR database is very confusing to me and the R package seems to feel the same way about it. The error message I am getting at this line is
"Error in c(matrixSet, list(xmatrix)) : could not find symbol "recursive" in environment of the generic function"
I googled it and found on your website that somebody told that it is caused because of a problem with Windows but that it works well under Linux. That is why I am trying so very hard to switch. Even people, who are experts in R could not find any mistake in my R code, which I only copy-pasted out of the vignette file.
It is hard to understand when even the sample code is not working. But if I cannot get it to work the novelty of my approach will be lost because I am trying to use a true similarity as a template based on which I want to calibrate the methods for determining the relative similarities between the time series plots for those genes because the order of relative similarities based on this TFBS package or the time series plots must be the same if we want to infer meaningful true similarities. Since there are many ways for determining similarities between time series plots I want to have a true guide consisting only of true similarities, which are encoded in some way in these time series plots too but the only way we can figure out how decode these similarities from these time series plots is if the relative order of similarities between genes is the same regardless of the dimensions of information based on which they are ranked.
Since I could not get this TFBS package to work for the past month already, I was looking for another way to get true guidance for similarities. This introduced me to an even more complex Bioconductor package, which is titled "GenomicFeatures"; See https://bioconductor.org/packages/release/bioc/html/GenomicFeatures.html. But this package looks even more intimidating because it seems to depend on more than 100 other Bioconductor packages. The installation of its dependencies goes way to fast for me to keep up with my eyes and therefore, I gave up on trying to figure out whether I even have all the its dependencies installed properly. I am just looking for a true sets of relative similarities in one dimension so I can use it as a model to calibrate the methods for relative similarity determination in another dimension because only then we can know the trues regardless of the dimensions we use for similarity ranking. Once that works we can use machine learning to come up with an algorithms that allows us to translate between both dimensions. And if we can quantify cellular responses in two unrelated seeming dimensions and get the same relative similarity rankings then we have become independent of a single dimension. This means that we have figured out a way to extract exactly the same information from the time series plots, which got initially decoded in them by the promoter specific TFBS distributions. Then we can define new dimensions, which are necessary and sufficient to reproducibly and ambiguously define other sub-aspects for translation for example, which we can then directly compare across different conditions. E.g. without needing any additional mRNA, the cell can decrease the distance between ribosomes, which run along a particular mRNA strand and translate it into proteins. And the nice thing is that we can see all the traces from this particular modification, which are sufficient for reconstructing the exact quantitative conditions and consequences of such kind of post-transnational modifications, which can be measured based on total mRNA and ribosomal coverage because every read = exactly one ribosome, which covered between 18-31 nt bps. The cell has about 10 options for changing translation without needing to change transcription but eventually its options are finite and given enough data can eventually be predicted and their impact on life can be quantified. Following this approach we should seen get to the point where we have highly predictable transparent cell response, which we can then manipulate more effectively in order to maximize benefits. So if you know of an easier way to rank genes or proteins based on some true similarities maybe I don't need to worry about these complex R packages anymore as long as I have one dimension of measurement based on which I can train another more distant dimension until both yield the same results and could then be considered as being understood, if not yet in mechanism, at least regarding their outcome. Does that make sense?
Because on the big Linux server I cannot figure out how to get R started.
If R exists on the server, and is correctly listed in the
$PATHvariable you can just launch R by typingRon the commandline. If you want to launch a script, you useRscript.In order to know whether R is installed on the machine however we REALLY need to see the output of a
which Rorcommand Rinput. It may also not exactly be called "R", so try lower case and so on. The output of these commands will be the exact path to the executable of R, and then we can move on.If you are looking to get the "graphical interface" from R going via the server you would need to use X-windows. This is another layer of complexity but you can't avoid it.
This is a tutorial on how to use X11 or X-Windows on windows.
I don't think so because a year ago I had a good tutor, who installed programs for me exactly on the same server but R and R Studio were not among them since back then I saw no urgent need to switch to Linux if R and R Studio worked well under Windows, where I am much more comfortable because there I can magnify and predict what my scripts will do.
Because if people find out that I don't know Linux , they would not let me use the server. So I must learn without anybody knowing that I don't know yet properly already. I have an idea of relating TFBS info with time series curves but this Bioconductor ToolsTRBS package cannot run under Windows. Once I read that last night I am panically searching for a way to move me R script on a Linux server to complete the task without waking suspicion that I almost would not have been able to complete it.
Don't worry about that. As a "regular" user you can't do any serious damage to the server. That would also not be a valid reason not to let anyone on the server. Unless you get an account how would you learn linux.
At least that was the reason why the dean did not give me permission to get an account on the powerful NSF server here at my school.
And here is an email I just received about not getting help because I know so little Linux. I know why I cannot trust people anywhere about needing more time to learn because this deficit is getting used against me over and over and over again. Just read this:
Thomas,
After reviewing the emails that you sent me, I do not believe that I will be able to help you with this project. It would benefit you to learn more Linux commands on your own before proceeding. It the current stage of this project I am concerned the time that will be expected of me and I will not be able to work full time on this. Furthermore I am VERY concerned that you are not technically supposed to be using the IU servers and therefore will not be able to help any further. Please look into other options for help.
Good luck,
These people aren't victimising you or using anything against you, they're simply stating that they cannot invest the amount of time it will seemingly take to get you up to speed. This is not unreasonable from what I can tell. You can learn all this stuff yourself in just a matter of days.
Ok, Tom, I don't mean to be a dick, but installation of R is about as well documented as it comes. If you're having trouble with the command line, then you're gonna have the same problem with R as it's a command line interface as well. And there's the inherent problem. You really cant hide that you don't know linux shell, but people aren't born with innate knowledge of bash shell, so you have to learn bash shell.
I finally got R to work under Linux and it was not even hard become somebody logged into my computer and told me what to type and the expected results.
Now I have 5 Linux commands saved in a word file, which are sufficient for running any R script under Linux and I learned them in less than an hour. Here they are:
1) Running R scripts in the background: R CMD BATCH /home/tom/tools/R/Rscripts/learning causality networks.R Output1Networks & 2) Shows which programs are all running. Like the task manager in Windows ps -x 3) ps –x 4) vi “output file name” shows what is normally printed into the console of R 5) To leave the vi editor, press Escape, then colon, then q and then !
Typing and reading takes often much longer than showing when prior experience is lacking.
I think this is the problem and written communication cannot overcome it:
When I presented, what I believe is my most powerful and far reaching but also most misunderstood proposal so far, at the 18th International ACM SIGACCESS Conference on Computers and Accessibility in Reno last week and also earlier this summer at the Research on Equity and Sustained Participation in Engineering, Computing, and Technology (RESPECT) Conference in Atlanta, I realized that when the necessary shared experiences for triggering the intended imaginations, sensations and memories are simply lacking, no string of words - and surprisingly not even images – can be powerful enough to instantly trigger the intended imaginations and sensation, which are absolutely essential for full understanding unless somebody first goes through the required experiences, sensations and frustrations, which seem to be irreplaceable prerequisites for forming the essential imaginations and sensations needed to truly understand in advance as part of a distinctly separate temporal process.
I also have 3 virtual boxes, on which I had R and R Studio installed. But a nasty network virus forced me to reformat my C drive and replace it with an image from May 2016. Now all the R and R Studio versions, which so many people spent so much time to install throughout last summer, are lost again. And since I have not used these virtual boxes for almost a year now because Windows was much more convenient to use instead of them, I have forgotten all the Linux admin passwords and hence I cannot reinstall R on these Oracle Virtual Boxes anymore either.
Actually, your suspicion was right. Even on the small server R Studio was already installed. I had somebody tutoring me remotely with screen sharing and we found it. But it was of no help because we still could not figure out how to get it to run. My tutor explained me that before I can run R Studio similar to what I am familiar with from Windows certain changes must be made in my profile and we could not figure out what these changes need to be. I hope we can figure it out tomorrow. The major problem is that we need to make sure that there is absolutely no chance that anybody here could become suspicious that I still have access to this particular server because they'd take it away immediately if they would suspect that I can still install programs in my home directory. That is why I cannot trust anybody here because the more people here would understand how I need to work in order to at least make some kind of progress the harder things would become for me here. They actually already took it away but don't know that I got help that managed to restore m access. Since they are so worried about me doing something bad to the computers they made me my own little mini-server, which basically consists of an old desktop with a few hard drives in it to give me some storage space for genomic analysis. But it is totally useless because on that old mini-server, which I have all for myself so that there is absolutely no risk for me to accidentally do something bad to somebody else files, I have absolutely no permissions, not even to install a new R package. For every new R package I want to try I need to email the technical director begging to install it for me. This can take up to two weeks or even longer. That is why I hardly ever use it. Sometimes people, who train me have some extra time unexpectedly. The we could try something I want to try for a long time but it must take place within a few hours rather than weeks because people get busy quickly with other things and their availability cannot be predicted 2 weeks in advance. I am sort of thinking that if they already gave me my own server, which is an old computer on the third floor, they could also take the risk for me to accidentally damage it, since it only has my files on it anyways. But if I had to depend on that server alone only things would be much worse than they already are. But complaining will make things only worse because the more I complain the more things will be taken away. And the problem is that if anybody here seriously starts focusing on proving that I am not good enough for this program this objective can very easily be achieved because even if I don't sleep for an entire week and only work to keep up with everyone else I cannot keep going on this fast paste for many months without taking a break. I can work none-stop for about up to 7 days with no sleep but then I cannot get up for about 3 days to catch up with the sleep deficit. The disability office here has already impressively proven this capability because they got me suspended for an entire year for having submitted a proposal about using remote training to make graduate school even a feasible option for me despite only having about 1/10th of the reading efficiency of sighted students by substituting reading with one on one tutoring. The disability office people here, especially the director felt threatened by my writings because they felt I'd intrude into their domain and expertise just because I can write much better then them and I know much better than any of them what helps and what does not. To achieve their objective they filed a title IV sexual embarrassment complaint against me since these are all women claiming that my true intention for having submitted my proposal to the AHEAD (Association of Higher Education and Disability Conference was to cause harm to them in some way I still cannot comprehend. And even my department chair urged me to immediately withdraw my proposal to avoid bigger trouble and I thought that was it. But although
NOBODY HERE
I don't mean to be a jerk, but we don't really need your life story. I feel bad for the predicament you are in but this site works best when answer threads are clear and concise so that people can refer back to them in months or maybe years time. Now to address some points: They simply cannot object to you installing software in your home directory. Any system administrator would know is this is absolutely harmless and a completely normal thing to do. Secondly if you want to run GUI Rstudio from the server, you will NEED an X11/XWindows client when you connect (PuTTY alone isn't enough for example). Then you have to hope against hope it's correctly configured on the server already otherwise you've got no chance, because it's freakin' difficult.
I withdrew instantly without complaining a bit and forwarded proof of withdrawal to my department chair it still gave them enough ammunition to use it as foundation for a title IV complaint accusing me that my true motivation for submission was of sexual nature. Everybody here knew that was absolutely bogus but nobody wanted to openly confront the very corrupt, fake and deceiving university disciplinary machinery here and advised me to accept this penalty and to refrain from filing a civil rights complaint for discrimination against my university because even those people supporting me told me that they could not take my side if I would bring a discrimination complaint against my university. And the problem was that I was not even allowed to defend myself because when I started soliciting experts for help to prove to these fake committees, which already had decided to suspend me before I came into the room to defend myself, because they construed out of me approaching experts for help that I had retaliated for the title IV complaint against me and therefore, instead of suspending me for only one semester, they effectively suspended me for three semesters. And then the dean and the vice chancellor, who masterminded this ironically told me: "If you had listened to our good advice and accepted the penalty without complaining about it, you would have only lost one instead of three semesters. And they kept changing the meanings of words and language in such a way that made no sense to anybody outside UALR but since inside UALR they are the one with despotic power the retro-active changes they made to language and meaning words words were the standards based on which any kind of ridiculous decision could be made. I felt like getting brainwashed because any common sense failed and everybody outside the university had a very hard time believing that such kinds of things can happen so easily and there did not seem to be anything that could be done to stop it. So any kind of enemy I create here will quickly grow clones, which can create problems much faster than I could ever solve them. Actually our program works like this: The first 3 semester’s students get financially supported by the graduate school while completing the coursework requirements. Then they need to find a mentor, who pays them for the remaining time. My problem was that nobody wanted to pay me since everyone knew that also I had a 3.98 GPA anybody else could do so much more work than me. So when this happened I went to the dean, but this was a good dean, and told him that I have no chance to continue. Then he asked me if I could survive on half of an assistantship, i.e. live on $750 instead of $1,500 per month. I told him yes and about four weeks later he was able to find some money for me from the EIT College instead of getting paid by my PI. Only then it was possible for me to find a PI and to stay in the program. But still, it took me 3 instead of 2 years to complete my master's. The initial idea was not to draw any more attention since they had already proven how easy it was for them to cause devastating damage in my life to come back as soon as the ban was over and to delay everything by a year because the enemies had already proven that they could easily flood me with so much bogus and counterproductive accusations and information that I had no more time to study even. The director of the disability office started to have secret meetings with my department chair about me trying to convince her that I should not be allowed and supported in continuing bioinformatics because I needed too much accommodations such as note-takes and readers when I still took full time classes and the disability office did not want to pay for that anymore. But these people are very fake. They always pretend they only have the best thing for me in mind even though they have absolutely no clue what this would be. But while I was gone, the good dean, who gave me the $750 per month so that I could stay in the program got promoted and became the dean of the entire graduate school because when he still gave me the money he was dean of the EIT collage.
And I knew that the other students got jealous because I was paid by the EIT College instead of my PI, which basically allowed me to dedicate all my time to studying. And the main problem here is that we don't even have enough money to pay all the normal PhD students. That is why they were upset that I had a secure funding base even though everything was taking me longer. Now while I was gone we had major budget cuts here at UALR and as of June 30th, 2017, my funding from the EIT College will be terminated. So I effectively only had funding for the first 10 months of my PhD program, which I started when I was no longer banned on August 16th, 2016, when I finally was allowed to return. So now instead of having 5 years to complete my PhD I am expected to finish in four years, which is completely impossible unless I can find people, who can help me to quickly get a publication based on which I could be given the degree. My adviser and I were trying this at the beginning and everyone else would have probably played along with it if it only had worked. But if I'd started to officially complain about it, this would only cause everyone here to unite against me and play along with those people, who want me to be gone anyways. Because here, it works like that: I am lucky that my PI, without glasses can see about as much as me. Therefore, he has never used the shortcomings of my low vision against me. He has given me tasks to complete but if I could not he gave me other tasks instead whereas most PIs would have quickly fired me if I cannot meet their expectations. But it was clear from the beginning that I would be responsible to find the resources for learning bioinformatics from the internet and people to teach me because nobody in my advisory committee knows much about bioinformatics. But I chose the people in my advisory committee not at all based on how much they knew in what I am expected to learn but based on how much I could trust them that they would not use any problems against me. That is much more valuable. Many professors told me that they would not be interested in me working for them even if they don't need to pay me because the time it takes me to learn new skills and to figure things out would not give them a good enough investment on their time. So actually - according to the student handbook I should have defended my proposal by now but since I still don't have a good results i had nothing I could defend. So at any time a single person, who wants to use this against me, will easily succeed, and it would look to every outside observer that I simply did not deserve to stay any longer in the program. I know this can happen any time now and it would immediately destroy me but it has not happened yet because at the moment nobody seems to focus on causing such kinds of problems as long as they are not aware that I am even here. I'll spend up to 7 days in my office at a time without meeting anybody in person, only through the internet, because ever since I got suspended I can no longer trust people and I don't even want anybody coming to me since the less people can understand me the less of a threat they can be for me. That is what I learned from the suspension. Sometimes my questions are too simple and basic because I cannot read properly and at times even my adviser is upset at me because I don't have a clear idea about his research. So people, especially when they are in control of essential resources without which I cannot function are much more a threat than a help because they would know about all the things I cannot do properly. This would make it too easy to create situations, which would in a fake but nevertheless convincing way show that I am not good enough to deserve to stay here. When I started at UALR we all were friends in the first semester, which was my hardest semester, but suddenly my friends from yesterday because my worst enemies of tomorrow. That why I don't even want to have friends anymore and because of the suspension trauma I did not make a single new friend after the ban against me had finally ended. I only have those friends, who I already had before I got suspended.
Seriously, none of this matters here. It in no way affects your bioinformatics problems. We've given you all the information you could possibly need. You've even confirmed R is already installed, which if you had just run
which Rlike one of the first comments told you to, you'd have answered this question a day ago...No it were 2 servers - actually five. One one of them I did not know that R was installed but of the IU servers it said in the web that they had R. So we tried the small server this evening about which I did not know. Bit even the person, who knows Linux and who trained me tonight could not tell me which changes I'd need to make to my profile so I can run the R we discovered today. Most of the people, with whom I have been collaborating only use Windows as they seem to feel as lost in Linux as I do. I have another kind of R, which looks like an IP address but with that one I cannot figure out which portions are taking place on my laptop and which portions are actually taking place on the server.
So everything is very tricky and can change at any time. Before I came here I was paid $32,000 per year at UTSW. But then I did not know yet how fake everything is and i filed a discrimination complaint against UTSW. But needless to say, I lost and got expelled. So if I start defending myself I will inevitably lose. A little bit like this is going on here too because the custodian does not like me and keeps complaining about me to the department chair. He vacuums everyone's office, except for mine. And in the night before I flew to the ERN conference in Washington last week I accidentally spilled some food on the floor of my office. But the custodian keeps discriminating against me by making me the only person, who he refuses to clean the office even though he cleans everyone else's office every morning. And he would not even let me use the vacuum to clean it myself. When I ask him he pretends not to hear me. So it happened last Friday morning when I was in Washington that the new EIT Dean decided to convince my department chair to inspect my office and found crumbs on the floor. The new EIT dean already wrote me a very official seriously written warning letter that if my office is not clean he'll have to take it away and I'd have not more space to study and talk on Skype to my tutors. But still, I cannot complain against him even though he is very obviously not doing his job. But nobody here forces him to if he does not feel like it. He is upset that over the weekend I drink so much energy drinks, snacks and canned food that on Monday morning the trashcan in the kitchen is full. Then he complains about me to the department chair so she keeps sending me warning emails But all I can do in this situation is to bag my old friends, who have left the university long time ago, to please come and bring a vacuum because else the dean will take my office away. He had already taken away the keys for almost 3 weeks during Christmas break during which I could not make any progress because in the small room I am renting there is simply not enough space to put a 50 inch screen.
Most people are strange and have caused so many problems that I seriously regret to have trusted them in the first place.