Skip to content

Getting Started

Pydre is a Python application, run from the command line.

pydre.run

The main entry point to run Pydre is pydre.run. This application that allows the user to analyze data using command line arguments.

The user must enter the path for the project file and data file in order to aggregate the data. The user has the option of specifying an output file name, to which the test results will be saved. If no output file name is given the output will save to "out.csv" by default. A brief description of the aforementioned arguments is offered below.

Command Line Arguments for pydre.run

  • Project File [-p]: The project file specifies a file for the regions of interest from which the data should be aggregated. It also specifies a list of metrics (the type of data) that will be aggregated from the regions of interest.

  • Data File [-d]: The data file contains the raw metrics obtained from the simulation. This argument supports wildcards.

  • Output File [-o]: After the script has executed, the output file will display the aggregated metrics from the regions of interests that were both specified in the project file. The output file will be saved in the same folder as the script.

  • Logger level [-l]: This defines the level the logger will print out. The default is 'warning'. Options include debug, info, warning, error, and critical.

Command Line Syntax: 

python -m pydre.run -p [project file path] -d [data file path] 
    -o [output file name] -l [warning level]

Example execution: 

python -m pydre.run -p C:\Users\pveith\Documents\pydre\docs\bioptics.json 
    -d C:\Users\pveith\Documents\bioptics\pydreDataSet\*.dat 
    -o bioptics.csv -l debug

Example data processing

Source files

We'll walk through the creation of a project file and running that project file on an example data file.

DatTime SimTime MediaTime LonAccel LatAccel Throttle Brake Gear Heading HeadwayDistance HeadwayTime Lane LaneOffset RoadOffset Steer Velocity XPos YPos GazeObj
1.1 9.1 5.2 0.014943 0.020915 0 0 3 0.116032 10000 486.5342 2 0.1 5.1 0.003261 20.55353 -6470.89 14.6 InstrumentCluster
2.1 10.1 6.2 0.017797 0.020037 0 0 3 0.116593 10000 486.5269 2 0.12 5.12 0.003104 20.55384 -6460 15.1 InstrumentCluster
3.1 11.1 7.2 0.019815 0.019082 9.814428 0 3 0.117187 10000 486.5185 2 0.15 5.15 0.002947 20.5542 -6449.11 14.6 None
4.1 12.1 8.2 0.02161 0.017984 9.814428 0 3 0.11781 10000 486.5088 2 0.2 5.2 0.002633 20.55461 -6438.21 14.65 None
5.1 13.1 9.2 0.023096 0.016835 0 4.3 3 0.118379 10000 486.4993 2 0.2 5.2 0.002319 20.55501 -6427.32 15.2 None

The example file is only five rows of a data file, where real data files are tens of thousands of lines. Also, this is synthetic data, for example purposes only, not real data from a driving simulator run.

Goals

Say you wanted to find the standard deviation of lane position (SDLP) for this data. Pydre comes with a standard deviation metric function, so you can use that. To do that, you would write a project file like so:

tutorial_project.toml
[metrics.SDLP]
function = "colSD"
var = "LanePosition"
The colSD function requires a parameter var that is the column name in the dat files to calculate the SD over.

After saving the project file, you can then run:

python -m pydre.run -p tutorial_project.toml -d Experimenter_S1_Tutorial_11002233.dat -o tutorial_out.csv

This will run the project commands on the specified data file and out the result in tutorial_out.csv. If you wanted to run on multiple data files, you could do -d *.dat or -d ../datafiles/tutorialdata/*.dat or something similar. Since we have no ROIs, the output csv will contain one row per input data file:

Subject Mode ScenarioName UniqueID ROI SDLP
S1 Experimenter Tutorial 11002233 0.04560701700396553

In this example, ROI is blank since we did not specify any ROIs. The metric column header is "SDLP", since we defined it under the category "metrics.SDLP" in the TOML project file.