malramsay64 / phd-thesis Goto Github PK
View Code? Open in Web Editor NEWMy completed PhD Thesis on Understanding Crystal Growth
My completed PhD Thesis on Understanding Crystal Growth
Are the results of the shape packing close enough to those using a LJ potential? Include
a comparison of the results.
Getting the energy of each crystal structure and the packing fraction as a way of comparing the validity of using packing fraction as a proxy for free energy.
Additionally it shouldn't hurt to include my updated results from the packing.
I have been using the code \exp{ei\theta}
and similar to denote the exponential, however this has issues in that it doesn't surround the text in any grouping symbol. Instead I should be using \exp(ei\theta)
so the parentheses show up.
Explain the choice of hoomd over other simulation programs.
This is the different scaling rules that have been developed. In writing them up, I will have a much better understanding of them, and can include some of the ideas within my own analysis.
The style of the dynamics figures still includes the small text, so this needs to be
updated to reflect the new style.
This would require some changes, and templating. Additionally all the images would need to be included as svg rather than pdf. By using pandoc most of the infrastructure is already in place, just going to need some high level structural things and relatively minor changes.
The initial structures which have been generated are rather important in the overall idea of the crystal melting chapter so it makes sense to have some validation of these methods and algorithms in the form of describing how they work.
This should include how a rotational relation and orientation order parameter are calculated from a quaternion, including an analysis of the performance of quaternions generally.
This should include how I know that we have indeed reached equilibrium. Noting it is a slow process for the low temperature configurations.
There are many errors within molecular dynamics calculations. This should document
It is really frustrating having to go from text to bibliography and back. To simplify this process there should be links within the bibliography back to the pages where the citations appear to make this process easier.
There are still a number of sections which just contain dot points. This is intended as a broader tracking issue of all the writing that is required with items ideally being checked off fairly regularly.
There are a range of theories of crystal growth, from the classical to other methods
The purpose of averaging over many angles is that we have an isotropic liquid, so any angle which we pick will be equivalent. This should be made explicit in the text.
While #54 introduced backlinks to the bibliography, the links are not styled particularly clearly. This should format them in a way which is clear and readable. Also making note that it is a link.
Some of the unicode symbols (namely a subsript m) are missing from whatever font is
being used in the svg file for the Altair figures. I need to check the font being used,
and it might be a good idea to match the font of the thesis.
There should be an image with the position of each of the atoms in the molecule, includign the virtual R atom. This will assist in the description of the molecule in the methods section.
The figure currently looks terrible
I need to be able to explain how I have calculated the melting point, including figures
detailing why.
This should include the equations for the calculations and how I am averaging the quantities and how I am calculating the mean values.
This also includes the relaxation quantities.
There is a lot more that needs to be written in the recommendations for computational reproducibility, taking information from the range of sources which I have cited.
This should add all writing checks as pre-commit hooks. This means that they shouldn't get through to the CI to fail there.
Which measurements went into determining the temperature of the spinodal, where they came from and why I have arrived at this number.
The lewis wahnstrom model is a simple system used to model the behavior of orthoterphenyl. This is going to be very similar to the types of analysis I am performing, and should be very useful as a comparison.
This provides the ability to cache a collection of the files. Since it takes so long to download from tectonic, caching the build files is a great idea.
I should start writing up the work I have done on the fluctuations, which will be addressing the why is this so difficult to melt.
Have a structure for the introduction which can be used to start writing
I have some dot points on this, however this is not yet written up in paragraphs
including figures.
There are many different quaternion distance methods described in @Huyng2009. There should be some description of why this particular method is correct.
Additionally the computational complexity should be discussed.
Include the figures of each crystal in the methods chapter
Much of the work I have done in the statistics of the dynamics calculations assumes that configurations are independent. This is rather different to the standard statistical methods which use a linear sequence of steps. This is mostly so the statistical methods which I use later can be considered valid.
It is fine having an understanding of simulation dynamics, however being able to relate my results to those of experimental systems is far more powerful.
Include how each of the relaxation times were calculated. Including the calculation of the actual value and the estimation of the error associated with that value.
By using a filter for pandoc to convert the svg images into pdfs, I am able to choose the output format depending on the document format, which solves many of the issues in #5
The only change should be the inclusion of an additional processing file, since it is currently not installable.
Explain how the numbers for these quantities were arrived at, including an estimation of the uncertainty in these values.
This is the value from the normalised melting rates, found as the parameter of fit. This value should be tabulated as a result.
I currently have a text diagram which should be converted to tikz for inclusion in the thesis.
Currently there is a component on the Crystal detection in the Methods chapter and in the Machine learning chapter. These should all be consolidated into the machine learning chapter which would be a more sensible place to put it.
Additionally the discussion on alternative methods, like the orientational order parameter will be useful in the crystallisation chapter where I will be looking at free energy.
Rather than having a job fail and have to look into why, this should include a style linter job and a compilation job.
There is a bug in the install command, using conda env update instead of create.
Furthermore, there is no documentation of the npm packages I am using for editing.
This should give a broad overview of each of my different projects, including how they relate to other projects and what they achieve.
How random numbers were used, in particular in the packing code. This should also include some descriptions of the algorithms and which are suitable pseudo random number generators like my talk https://github.com/malramsay64/random_number_generators.
This should include the conditions for initialising the dynamics simulation including the minimisation and other steps
Most of the measurements of viscosity use the unit of measurement Poise, which should be discussed in the introduction. In particular information about the viscosity of glasses, particularly since this will probably include the Angell plot.
Collect and analyse the crystal melting results
The error calculations 2.2.2 are rather sparse and could do with a lot more explanation of the process.
I want to use git-lint to lint my commit messages ensuring that there are a standard set of messages. These would include at least:
Collect results from ML Crystals project
A declarative, efficient, and flexible JavaScript library for building user interfaces.
๐ Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. ๐๐๐
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google โค๏ธ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.