Mlflow not running on machine - r

Please I am trying to run mlflow code in R after having installed it. However, after loading the library with library(mlflow) and I run mlflow_log_params("foo",42) I get the error message below printed in my console:
Error in rethrow_call(c_processx_exec, command, c(command, args), pty, :
Command 'C:/Users/IFEANYI/AppData/Local/r-miniconda/envs/r-mlflow-1.19.0/mlflow' not found #win/processx.c:982 (processx_exec)
I also get the same error message when I run mlflow_ui(). Please was there something I ought to have done during installation failure of which is affecting its functionality? Do I need to install and load the processx library in order for mlflow to run on my Windows10 machine? I really hope I can get advice to help me because I want to use mlflow in my machine learning projects. Thanks in advance of your generous help.

The error should disappear when setting MLFLOW_BIN system variable (Windows) to mlflow cli executable : "....conda\envs\r-mlflow-1.24.0\Scripts\mlflow.exe".
If it works please mark the problem as resolved.
Unfortunately, you will get next error "Error in wait_for(function() mlflow_rest("experiments", "list", client = client)" for which I cannot find the solution

Related

How can I get past this 'SSL connect error' when using jsonlite::fromJSON in R?

Issue
I get the following error whenever I run
all_companies <- jsonlite::fromJSON("https://www.sec.gov/files/company_tickers_exchange.json")
Warning: URL 'https://www.sec.gov/files/company_tickers_exchange.json': status was 'SSL connect error'Error in open.connection(con, "rb") :
cannot open the connection to 'https://www.sec.gov/files/company_tickers_exchange.json'
Thank you in advance for any help!
What I've tried
I am trying to run this on a new work computer, it runs fine on my old work computer but the new one gives me this error.
It seems the difference causing this issue is that I am using {jsonlite} 1.8.2 on my new computer and 1.8.0 on my old computer. Deep in the definition of the fromJSON() function, the working version seems to use curl::curl() to establish a connection whereas the not-working version uses base::url().
The following example line from the help documentation runs fine on the new computer with version 1.8.2
data1 <- fromJSON("https://api.github.com/users/hadley/orgs")
I can access the JSON file I am trying to read on an internet browser.
I do not have permission to install Rtools on my computer to be able to compile an older version of {jsonlite}.

how can fix this issue "Error in h2o.upload_model() : could not find function "h2o.upload_model"?

I have installed h2o package(in R from RStudio console). Post h2o.init() I am trying to use the built in function upload_model()/upload_mojo() but I am getting following error.
h2o.upload_mojo()
Error in h2o.upload_mojo() : could not find function "h2o.upload_mojo"
h2o.upload_model()
Error in h2o.upload_model() : could not find function "h2o.upload_model"
I found work around to resolve this issue. Please find below the steps I followed:
Remove the package using: remove.packages("h2o")
Quit the current session and launch the new one.
Move out lock file for h2o from path where package was installed mostly under R with file name like - 00LOCK-h2o
Install new/latest version of package via RStudio console using install.packages()
It should now resolve this issue.

Shinyloadtest: Error: Please upgrade your pandoc version to be at least v2.2

I am trying to run results from recording.log, as instructed from https://rstudio.github.io/shinyloadtest/ I am running below command to see in browser view:
shinyloadtest::shinyloadtest_report(df, "run1.html")
but throwing an error :
to resolve this, when I am trying install pandoc, throwing below error
what I am doing wrong in this?
can any one please help me on this!
My Issue was resolved, I have used one command in R Studio that can overpass pandoc issue:
shinyloadtest_report(df,"run1.html",self_contained=F)

Problems with R Request price feed through Bloomberg API

Following the steps as outlined below:
install Java APIv3 from the Bloomberg terminal (by typing WAPI into the command bar). Once installed connect it to R using :install.packages("Rbbg", repos = "http://r.findata.org") and conn <- blpConnect(log.level = "finest"). Finally, to extract share price information you use bdp(conn,securities,function)
I get an error when trying to connect that gives me the following message:
Error in .jnew("org/findata/blpwrapper/Connection", java.log.level) :
org.findata.blpwrapper.WrapperException: Session not started because: CONNECTION_FAILURE
Any advice how to resolve this would be very appreciated.
Please try migrating to Rblpapi. This is more modern equivalent and can be found on CRAN (install.packages("Rblpapi")) or github (https://github.com/Rblp/Rblpapi).

error while loading shared libraries: /usr/lib/libstdc++.so.6: file too short

I'm running this on CentOS 6.8 (Final)
I'm trying to setup a Arma 3 Wasteland server for me and some friends.
But when I try to run it it gives the following error:
./arma3server: /usr/lib/libstdc++.so.6: version `GLIBCXX_3.4.15' not found (required by ./arma3server)
I'm not an advanced linux user, I'm more of a noob, I tried googling but it doesn't work, is there anyone that could help me fix this? Just give me a step by step guide of what I have to do please. If you need more info then please ask I'd be happy to give more information.
After trying to fix it I now get this error:
./arma3server: error while loading shared libraries: /usr/lib/libstdc++.so.6: file too short

Resources