EMV_tool

EMV stands for "Experimental Metrics Visualization". EMV_tool is a simple tool for visualizing the metrics of a repetitive experiment. It is a living repository and new commits will appear when new updates occur.

All the experimental iterations should be placed in a single numpy array with the following structure. The height of this array represents the several repetitions of the experiment; each row corresponds to a single repetition, the width represents the time points (x-axis) of the experiment, while the depth depicts the different metrics used for the evaluation of the experiment.

Sample input numpy array:

Shape: (100, 10, 3) -> The experiment has 100 timepoints, it has been conducted 10 times and 3 metrics were used for the evaluation.

Configuration parameters:

General parameters:

• med_color: color of the line that corresponds to median value
• var_color: color of the deviation area around median
• xlabel: label for x-axis
• ylabel: label for y-axis
• font-size: font size for labels of the figure

fig_args = {
  "med_color": '#219ebc',
  "var_color": '#219ebc',
  "xlabel": 'Time Points',
  "ylabel": 'Metric Name',
  "font_size": 14
}

Simulation parameters:

• num_of_timepoints: The number of experiment x-axis points
• num_of_iterations: The number of experiment repetitions
• num_of_metrics: The number of experiment metrics
• metric_id: Metric id for the sample metrics
• window: Variation number

gen_args = {
  "num_of_timepoints": 100, 
  "num_of_iterations": 20,  
  "num_of_metrics": 3,      
  "metric_id": 2,           
  "window": 0.8 
}

Additional libraries

• numpy
• scipy
• matplotlib

Installation

The EMV_tool can be cloned from here or it can be installed using Python pip tool

• Option 1: Clone the repository and see the exmple case in test_bed.py file
• Option 2:
  • Install tool using pip3 install git+https://github.com/itzortzis/EMV_tool.git
  • Import the needed components from emv import utils or/and from emv import sample_data_utils as sdu

Sample outputs