I'm building a R package for binary classification and I'm using opencpu to host it. Currently I've saved the h5 file as .RData file(serialized), which is then loaded in the environment using the .onLoad() function in R. This enables the R script to use the environment variable to load keras model using keras::unserialized_model().
I've tried directly using keras::load_model_hdf5() in the code, but after building and deploying on opencpu, when I try to hit the prediction API, I get error
ioerror: unable to open file (unable to open file: name = '/home/modelfile_26feb.h5', errno = 13, error message = 'permission denied', flags = 0, o_flags = 0)
I have changed permission for the file(777) and even the groups but still getting the error.
I even tried putting the file in inst/extdata folder so that it gets in the package but still same error.
Can anyone help on this, or suggest some alternative to load the h5 model directly?
Which OS does OpenCPU run on? Why does it try to write in /home/, this is very unusual? The best solution is to adapt your code to write in getwd() or tempdir(). Even better is to store data in a local database or redis server and let R read it from there, so you don't need disk access at all.
If you run on Ubuntu Server, reading from /home/ is not permitted by default. If you want to allow this, you need to add apparmor rules, see section 3.5 of the server manual.
Some relevant topics from the opencpu mailing list:
write in home dir: https://groups.google.com/d/msg/opencpu/5vRvgSKY-qE/4xMzZCGJBAAJ
keras in opencpu: https://groups.google.com/d/msg/opencpu/HhRzFVVFdaA/n5Nu1sxyFgAJ
write tmp folder: https://groups.google.com/d/msg/opencpu/Y1tYhaQUzwU/ubSEd_CDCgAJ
Related
I am learning R and followed the instructions to program R using Visual Studio Code. I then tried to run the following line of code to learn how to read data.
dat <- read.table("d.data")
View(dat)
where d.data is a data file. I received the following error:
cannot open file '...\.vscode-R/request.log': Permission denied.
I tried using the "Give Access To" command from right-clicking the file in File Explorer, however, I don't think it did anything. How do I grant the program/terminal permission to open the file? It may be significant to note that running the same commands using the radian console works without any issues (I get the data outputted in a separate window).
I found a workaround for this issue by adjusting these settings in VSCode:
"r.alwaysUseActiveTerminal": true,
"r.bracketedPaste": true,
Then, by calling radian in an opened cmd terminal, I was able to load everything without any issues
When running R CMD check I get the following note:
checking for future file timestamps ... NOTE
unable to verify current time
I have seen this discussed here, but I am not sure which files it is checking for timestamps, so I'm not sure which files I should look at. This happens locally on my windows and remotely on different systems (using github actions).
Take a look at https://svn.r-project.org/R/trunk/src/library/tools/R/check.R
The check command relies on an external web resource:
now <- tryCatch({
foo <- suppressWarnings(readLines("http://worldclockapi.com/api/json/utc/now",
warn = FALSE))
This resource http://worldclockapi.com/ is currently not available.
Hence the following happens (see same package source):
if (is.na(now)) {
any <- TRUE
noteLog(Log, "unable to verify current time")
See also references:
https://community.rstudio.com/t/r-devel-r-cmd-check-failing-because-of-time-unable-to-verify-current-time/25589
So, unfortunately this requires a fix in the check function by the R development team ... or the web-resource coming online again.
To add to qasta's answer, you can silence this check by setting the _R_CHECK_SYSTEM_CLOCK_ environment variable to zero e.g Sys.setenv('_R_CHECK_SYSTEM_CLOCK_' = 0)
To silence this in a persistent manner, you can set this environment variable on R startup. One way to do so is through the .Renviron file, in the following manner:
install.packages("usethis") (If not installed already)
usethis::edit_r_environ()
Add _R_CHECK_SYSTEM_CLOCK_=0 to the file
Save, close file, restart R
I tried to load data to my R working directory and receive this error:
Error: bad restore file magic number (file may be corrupted) -- no data loaded
In addition: Warning message:
file ‘classize.RData’ has magic number 'RDX3'
Use of save versions prior to 2 is deprecated
I googled it and tried many options, unsuccessfully.
My Rstudio version is: 1.2.5033 (The error was happening before updating as well)
I create a new project, in the new directory, I put the data file
The data file is "classize.RData"
I have another alternative which is "classize.RDS" with the sugesstion to use readRDS(file = "classize.RDS"). When using this command, I receive that error:
cannot read workspace version 3 written by R 3.6.1; need R 3.5.0 or newer
This is in the context of a statistical course at university and my teacher assistant is unable to help me out, and whitout resolving this issue, I cannot move forward in the resolution of the needed exrecices. So please, couly you help me resolve that problem.
ps: all the students have access to the same data, It's just for me that it's not working, therefore the file should not be corrupted.
I want to deploy a basic trained R model as a webservice to AzureML. Similar to what is done here:
http://www.r-bloggers.com/deploying-a-car-price-model-using-r-and-azureml/
Since that post the publishWebService function in the R AzureML package was has changed it now requires me to have a workspace object as first parameter thus my R code looks as follows:
library(MASS)
library(AzureML)
PredictionModel = lm( medv ~ lstat , data = Boston )
PricePredFunktion = function(percent)
{return(predict(PredictionModel, data.frame(lstat =percent)))}
myWsID = "<my Workspace ID>"
myAuth = "<my Authorization code"
ws = workspace(myWsID, myAuth, api_endpoint = "https://studio.azureml.net/", .validate = TRUE)
# publish the R function to AzureML
PricePredService = publishWebService(
ws,
"PricePredFunktion",
"PricePredOnline",
list("lstat" = "float"),
list("mdev" = "float"),
myWsID,
myAuth
)
But every time I execute the code I get the following error:
Error in publishWebService(ws, "PricePredFunktion", "PricePredOnline", :
Requires external zip utility. Please install zip, ensure it's on your path and try again.
I tried installing programs that handle zip files (like 7zip) on my machine as well as calling the utils library in R which allows R to directly interact with zip files. But I couldn't get rid of the error.
I also found the R package code that is throwing the error, it is on line 154 on this page:
https://github.com/RevolutionAnalytics/AzureML/blob/master/R/internal.R
but it didn't help me in figuring out what to do.
Thanks in advance for any Help!
The Azure Machine Learning API requires the payload to be zipped, which is why the package insists on the zip utility being installed. (This is an unfortunate situation, and hopefully we can find a way in future to include a zip with the package.)
It is unlikely that you will ever encounter this situation on Linux, since most (all?) Linux distributions includes a zip utility.
Thus, on Windows, you have to do the following procedure once:
Install a zip utility (RTools has one and this works)
Ensure the zip is on your path
Restart R – this is important, otherwise R will not recognize the changed path
Upon completion, the litmus test is if R can see your zip. To do this, try:
Sys.which("zip")
You should get a result similar to this:
zip
"C:\\Rtools\\R-3.1\\bin\\zip.exe"
In other words, R should recognize the installation path.
On previous occasions when people told me this didn’t work, it was always because they thought they had a zip in the path, but it turned out they didn’t.
One last comment: installing 7zip may not work. The reason is that 7zip contains a utility called 7zip, but R will only look for a utility called zip.
I saw this link earlier but the additional clarification which made my code not work was
1. Address and Path of Rtools was not as straigt forward
2. You need to Reboot R
With regards to the address - always look where it was installed . I also used this code to set the path and ALWAYS ADD ZIP at the end
##Rtools.bin="C:\\Users\\User_2\\R-Portable\\Rtools\\bin"
Rtools.bin="C:\\Rtools\\bin\\zip"
sys.path = Sys.getenv("PATH")
if (Sys.which("zip") == "" ) {
system(paste("setx PATH \"", Rtools.bin, ";", sys.path, "\"", sep = ""))
}
Sys.which("zip")
you should get a return of
" C:\\RTools|\bin\zip"
From looking at Andrie's comment here: https://github.com/RevolutionAnalytics/AzureML/commit/9cf2c5c59f1f82b874dc7fdb1f9439b11ab60f40
Implies we can just download RTools and be done with it.
Download RTools from:
https://cran.r-project.org/bin/windows/Rtools/
During installation select the check box to modify the PATH
At first it didn't work. I then tried R32bit, and that seemed to work. Then R64 bit started working again. Honestly, not sure if I did something in the middle to make it work. Only takes a few minutes so worth a punt.
Try the following
-Download the Rtools file which usually contains the zip utility.
-Copy all the files in the "bin" folder of "Rtools"
-Paste them in "~/RStudio/bin/x64" folder
I have created the R script environment (used this command to create it "create script environment RSCRIPT command '/usr/local/R/bin/Rscript --vanilla --slave'") and tried running the one R script but it fails with the below error message.
ERROR: RS 10 S 332659 R 31A004F LO:Script stderr: external script vfork child: No such file or directory
Is it because of the below line which i am using in the script ?
mydata <- read.csv(file=file("stdin"), header=TRUE)
if (nrow(mydata) > 0){
I am not sure what is it expecting.
I have one more questions to ask.
1) do we need to install the R package on our unix box ? if not then the kognitio package has it
I suspect the problem here is that you have not installed the R environment on ALL the database nodes in your system - it must be installed on every DB node involved in processing (as explained in chapter 10 of the Kognitio Guide which you can download from http://www.kognitio.com/forums/viewtopic.php?t=3) or you will see errors like "external script vfork child: No such file or directory".
You would normally use a remote deployment tool (e.g. HP's RDP) to ensure the installation was identical on all DB nodes. Alternatively, you can leverage the Kognitio wxsync tool to synchronise files across nodes.
Section 10.6 of the Kognitio Guide also explains how to constrain which DB nodes are involved in processing - this is appropriate if your script environment should not run on all nodes for some reason (e.g. it has an expensive per-node/per-core licence). That does not seem appropriate for using R though.