资源算法yellowfin-r-roll

yellowfin-r-roll

2020-02-20 | |  36 |   0 |   0

yellowfin-r-roll

An advanced analytics function for Yellowfin to incorporate R scripts into report generation

This project is an advanced function integration into Yellowfin's (www.yellowfinbi.com) report development interface so that R scripts can be called on. Prior to applying this function to a dataset in Yellowfin you will need the R-script in a location on your Yellowfin application server. The parameter to populate the location of the R script should be filled with an absolute filesystem path, such as /Users/joe/model.r, where Yellowfin has the authority to read the script in that path.

The script may assume that a dataframe (which will be populated by the integrated function) contains a dataframe called yf_dataset. That dataframe can have up to 5 fields (parameter0, parameter1, ...,parameter4) which represent the 5 features for each record that the model can be applied to. The assumption is that the model is applied to each of the feature sets and yields a value. That value can be written then as a list to yf_dataset$result.

In Yellowfin all 5 parameters must point to a column in the dataset. Any unused parameters will simply be ignored. The reason for this is that there is no way currently to have a parameter that points to a column set to empty. This is being discussed with the Yellowfin team and may be updated in the future.

yf_dataset$result is the value that will appear in the resulting column in the Yellowfin report. If your R script uses CRAN packages, you will need to follow the instructions on www.renjin.org for integrating those CRAN packages into your local environment.

Not that the slower the R-script is, the slower the report will be to generate.

To simply download and run this project, download the jar files from the latest release. Copy those files into your Yellowfin web application library ([Yellowfin Install Directory ]/appserver/webapps/ROOT/WEB-INF/lib) and restart Yellowfin. You should see the additional functions available once you build a new report.

To build this project you will need to include the libraries provided in the /deps folder in your build path. You can also get these online through the www.renjin.org project, as well as the log4j repository. You will also need to reference the two jar files below in your Yellowfin web application library : [Yellowfin Install Directory ]/appserver/webapps/ROOT/WEB-INF/lib

  • i4-core.jar

  • i4-mi.jar

Once you have built the jar from the source code, copy the resulting jar as well as the contents of the deps folder into your Yellowfin web application library (see path above). Restart Yellowfin, and when you build a report and add an advanced function (more here : http://wiki.yellowfin.com.au/display/USER71/Advanced+Functions).


上一篇:yellowfin-http2

下一篇:yellowfin_sample_connector

用户评价
全部评价

热门资源

  • seetafaceJNI

    项目介绍 基于中科院seetaface2进行封装的JAVA...

  • spark-corenlp

    This package wraps Stanford CoreNLP annotators ...

  • Keras-ResNeXt

    Keras ResNeXt Implementation of ResNeXt models...

  • capsnet-with-caps...

    CapsNet with capsule-wise convolution Project ...

  • shih-styletransfer

    shih-styletransfer Code from Style Transfer ...