Usage

Access library

On balfrin simply add the following line to your ~/.bashrc file

export PATH="/store/msrad/utils/anaconda3-wolfensb/bin:$PATH"

Reload the session and then simply run the command

source activate radardb

Which will load the appropriate anaconda virtualenv, you can then directly use the command line tools (Interface, Database command-line tool, Machine-learning command-line tool, QPE command-line tools) in the standard console and/or import the library within python with

import rainforest

Note that on balfrin the radar database is stored in /store/msrad/radar/radar_database/

Keeping database up-to-date

Note

In the following, we expect that you are on balfrin you want to update the database but keep its current structure (same variables, same table names, same configuration). If you want something more fancy please read the doc carefully and decide by yourself.

To keep the database up-to-date you need to update the three tables gauge, radar and reference, always starting with the first one gauge as it is a prerequisite for the others. You can do this either by using the command-line Interface or by using the command db_populate.

db_populate -t gauge -s <start time in YYYYddmmHHMM format> -  <end time in YYYYddmmHHMM format>

db_populate -t reference
db_populate -t radar

Keeping randomForest model up-to-date

Note

In the following, we expect that you are on balfrin and want to update only the RFO RF model in its default configuration and covering the whole timerange of the database. If you want something more fancy please read the doc carefully and decide by yourself.

Once you have updated the database you can use the command rf_compute.

rf_train -m 'RFO'