MLabsPy

I origonally started this project while working for an in-house AV company, mostly as a Power Distribution Technicain, Rigger and AV technician, in my speare time. It was only a basic script that ploted 3 graphs, made some HTML files with plotly and had to be run from the console.

This module cleans and prepares machine generated Power usage data from a Motion Labs PM-XXX generated csv file using NumPy, so the data can be used for data analysis and graphical representation.
Jupyter notebooks were used for its ability to take notes and analyze and plot the data with matplotlib or plotly.

The data by itself is usless without context, this is the starting block to prepair the data before it's ploted or inserted into a Database. You want to tell a story of how the service was used, use the data to make predictions for possible up sales.

PM-XXX Series Power Meter Logger

Details for the local Data logging format and how to import into excel is discussed in the manual (pg 23, 8. Local Data Logging):

https://www.motionlabs.com/wp-content/uploads/PM-XXX-UM.pdf

Excel is slow with large datasets (45 x 7000), unless you have a good workstation. This python program (numpy) is much faster in my opinion. Jupyter Notebooks has been very helpful in developing this program and Spyder. The plots I made with excel in the past didn't look very large nor detailed. matplotlib does a good job marking larger plot images, plotly made some nice HTML/Js graphs with the ablity to zoom in and look at the data in ways you can't with excel or numbers

Prerequisets

• numpy
• scipy
• matplotlib (plot test)
• jupyter-notebooks (optional)

How-To Use

Bellow is all the information about the columns in the log file and the tag names to access the data after it has been parsed. Many of the tags have been shortened from the origonal one. I have also added tags for plots that have multiple lines plotted.

plotData returns a data structure like:

{
    'x': [1,2,3,4,5],
    'stats': {
        'min': 0.,
        'max': 10.,
        'median': 0.,
        'mean': 0.,
        'mode': [ 0, 1 ]
    },
    'plot': [ {  
        'dataset': [ 1.,5.6,7.4,6.,2. ],
        'tag': "float test",
        'color': 'red'
    }],
}

Using the MLabsPy.Plot class to plot the data

import MLabsPy as m

file = './test/test.csv'
ML = m.MLabsPy(file)

# Ploting
p = ML.plotData('v208')
plot = m.Plot(x)
plot.go(p)

Using the MLabsPy.MLabsPy class to get CSV data

getCSVdata() returns an array if CSV data, for file writting or pandas. this data has been cleaned, so it can be easily imported in excel after making the CSV file. the first value is a time stamp,the rest are floats.

import MLabsPy as m

file = './test/test.csv'
ML = m.MLabsPy(file)

CSV = ML.getCSVdata('v208')

More information in Jupyter Notebooks

I have created a jupyter notebook for this class to help you see the ploting in action. all the access tags are listed there along with some more information about the data being logged.