dhhagan / opcsim

opcsim is a python package for simulating low-cost optical particle sensors

Hi, I'm David! 👋

I'm a software engineer with a background in chemical engineering and aerosol chemisty and physics. I'm currently the CEO of QuantAQ - we build low-cost, professional-grade air quality sensors to supplement existing regulatory monitoring networks.

whoami

• I'm originally from the Bay Area (Napa, CA) and currently reside in Cambridge, MA
• I completed undergrad degrees in Chemical Physics (Hendrix College, AR) and Chemical Engineering (WashU, MO), and a PhD in Atmospheric Chemistry and Physics (MIT, MA)
• I love baseball (Go Giants!)

🏡 Location

I currently reside in Cambridge, MA and can often be found in one of the many coffee shops around (Broadsheet and Gracenote are favorites).

Previously, I've lived in:

• Napa, CA
• Conway, AR
• Thousand Oaks, CA
• St. Louis, MO
• Ann Arbor, MI
• Cambridge/Somerville, MA

I also spent a fairly significant chunk of time in India (mostly Delhi, Mumbai) during my PhD, though in 2-4 week chunks.

📱 Contact

It is easiest to reach me via email at the address listed on my profile. For air quality-related inquiries, you can also reach out to QuantAQ directly at [email protected].

code

I write a lot of software. I mostly work in Python and C++ and spend most of my time these days across a number of various areas including front-end, back-end, architecture, devops, and embedded software. In addition to writing code, I spend a good chunk of time designing the electrical and mechanical components that comprise QuantAQ's product line.

opcsim's Issues

Add an example of hygroscopic growth to the gallery

Requirements

Pandoc must be installed for making notebooks to rst

*left* kwarg to bar plots are being deprecated in favor of 'x'

Create a new OPC Class

Create a totally new class that is actually relevant. It should have "calibrate", "predict", and maybe other methods.

Re-write documentation for new OPC class

RI needs to be adjusted for water content (i.e. RH)

Verify the non-linear differences are real

Calculate where the midpoint in each bin would be binned given an RH (and corresponding RI) difference.

Calibration Plots

Make it easy to build a Mie theory calibration plot with the bin boundaries added

Mention that the diameter we use in each bin is the geometric mean

Create helper function to allow people to create complex materials

Do we actually need this?

Edit contributing instructions to be up-to-date

Add badges for code coverage and pypi version

Remove h20 from the list of RI options

Re-define an Optical Particle Sensor

Break it down into OPC's, Photometers, and Nephelometers. Discuss how the signals differ, give examples of common ones, and talk about their pro's and con's. Also, show some literature values.

Figure out why the OPC isn't reporting large particles

Is this just because the RI has dropped so much?

Add sample distributions

Add in the sample distributions from S+P

Ticks can be incorrect in scalar format

Specifically, '0.001' is showing up as '0.00'

Add examples to docs of new OPC class

Add kwargs for kappa and density to the individual modes of an aerosol distribution

Import error (equations)

installation isntructions

jon had some errors installing on windows without github installed...everything worked fine if he downloaded, unzipped, and installed...

Bug with evaluation of OPC with distribution

There seems to be a bug associated with choosing between simple and subint methods.

Fit Cscat to Dp

• Linear Fit
• Multi-line fit with breaks between non-monotonically increasing points
• interpolate across non-monotonically increasing points

Installation doc error for pip install through git

should be from https not git

Add support for RI for all distributions and modes

Plotting with_modes=True should use different colors

Make sure units on Mass conversions are correct

It may have been left in ug/cm3 rather than ug/m3

Edit tutorial for distribution to include RI and RH

Fix gallery with examples using new OPC class

Warnings during sphinx-autodoc

/usr/local/lib/python3.6/site-packages/matplotlib/figure.py:1999: UserWarning: This figure includes Axes that are not compatible with tight_layout, so results might be incorrect.

Pertains to figure ten_bin_opc.py

Add a new Nephelometer class

Take a look at how much we are over or underestimating each bin

Based on the difference between the fit curve and the actual value

Rewrite docs sections on aerosol distributions to reflect june 2019 changes

Support Mutli-Modal Fits

Support fitting real data with multiple modes.

Tutorial for new OPC class

Fix the plots

Computations for Nephelometers

For nephelometers, create the "calibrate" method which should essentially correlate the mass of a distribution to the Qscat. Also, calculate the bscat

Does the scattering angle make a significant difference in our calculations?

What happens if we plot scattering at a single angle vs the sum of a range of angles for a nephelometer? Does it really make a big difference? Or is it mostly a sensitivity/signal thing?

Add support for RH for all distributions

Update the project features

Build a way to generate histograms of cscat values

Compute bins from aerosol distribution

From an aerosol distribution, generate a cscat distribution and then assign each particle to a bin

Add ability to convert to mass-weighted things

Convert weight and base checks to decorators

Matplotlib Deprecation Warning: Bar plot

Error Trace:

MatplotlibDeprecationWarning: The *left* kwarg to `bar` is deprecated u
se *x* instead. Support for *left* will be removed in Matplotlib 3.0

Beginning with Matplotlib3.0, we need to address the bar plot API changes.

THIS DOESNT WORK FOR ME AND IT'S NOT YOUR FAULT

change colors on plots more easily

Scoring Algorithms

WARNING: Title level inconsistent

Appears during autodoc for file tutorial/distribution.rst

Better installation instructions

Add the installation instructions for pypi on this page.