Scrapbook (Posts about advice-tips)

Preparing Presentations

author — Mon, 12 Aug 2024 14:40:07 GMT

Getting Started

Before starting anything, ask yourself "Who is my audience and how much time will I have to present?" A 12-minute presentation at AIChE in the CRE division will likely look different compared to a 30-minute G.R.A.S.S. talk to the general audience that is the department.
See the Getting Started section in the Preparing Manuscripts page. If you haven't written a manuscript for this project yet, it's important to ask yourself these questions beforehand, as they serve as the foundation of the presentation. The way you answer these questions should be dictated by the assumed background knowledge your audience has (e.g., you don't have to describe what Monte Carlo is at a computational session).
At this point, you should section your work. Usually the standard separation of intro/background/motivation, methods, results/discussion, and conclusions/future work is sufficient, but you may need sub-segments in methods or results depending on whether the presentation is methods- or results-focused. For each segment/sub-segment, ask yourself "What is the overarching takeaway point for this section?"
- If you have a manuscript prepared (as with qualifying exams), read through the paper and write down what you think the main takeaway point is for each section.
- The segments shouldn't be the same lengths. Depending on time constraints, some sections may need to be shortened to allow for other sections to shine. A good rule of thumb is to have the intro/motivation only be 2–4 minutes so you can spend the majority of the time talking about your research.
With the key takeaways written, ask yourself "How do we go from one takeaway to another?" Essentially, what is the general story you want to tell? Keep in mind that the story you're crafting doesn't need to be told chronologically — this is usually the case for the results sections. If the analyses were performed A → B → C but a more compelling narrative can be told going A → C → B, then that's fine.
Only at this point should you ask "What figures, analysis background, system descriptions, simulation flowcharts, etc. do I need to ensure the takeaways are well understood by the audience?" (i.e., how can I make sure the audience appreciates this outcome?)
- Assuming you don't have some from the manuscript, when making these figures it may be easier to roughly sketch them on paper and see if they're clear. Ask a labmate if it makes sense to them without prompting them with what they're trying to understand.
At this point, the entire skeleton of the presentation should be done and it should be relatively clear how to make the slides or how to structure the poster.

Tips

Most importantly, always try to make sure your audience leaves feeling they learned something.
There are two ways to make slides, having slide animations animations (see Anne's slides) or having many slides that progressively build off each other (see Joseph's slides). Peng usually prefers the latter as it's easier for him to use if he needs to use your slides for a presentation. To do this effectively, make all the "main slides" with all the conclusions, figures, etc. in them, then progressively delete and hide sections. This will make sure that no figures/textboxes are being accidentally shifted. If you want, keep a copy with just the main slides so you can easily look back at the presentation.
Each "main slide" should roughly be about 1-1.5 minutes.
Often times, it's impossible to show every analysis you've done. This is usually for the best, as you can have more in-depth discussion on the analyses you are presenting.
For complex methods or analysis, it's usually better to explain and show things step-by-step rather than showing everything at once (time permitting).
Similarly, for graphs, try to explain what the x-axis is describing, what the y-axis is describing, and generally how to read the graph. Just showing a graph and saying "This is the conclusion you should be drawing" is a data dump and will leave the audience unfulfilled.
For slides with conclusions, write out the conclusion and make it clear that it is the key takeaway (either make it bold or put it in a box). This may seem unnecessary if you're saying the conclusion aloud, but ensuring that the audience is on the same page as you is always for the best.
For acknowledgements, computational resources (NERSC, ACCESS, etc.) will usually have approved images on their websites that they want you to use.
It helps to have a loose script for each slide. This ensures you know the talking points and is a good safeguard for when you stumble and need to recover, while also making sure the talk doesn't feel "too scripted."
Slide titles should be descriptive. Don't use something vague like "Analysis" as a title — be more specific, such as "Quantitative Analysis of Structure."

Talks

Ten simple rules for making good oral presentations (Philip E. Bourne, 2007)

Effective Presentations—A Must (Craig J. Hawker)

How to give a good research talk (Simon Peyton Jones, John Hughes, John Launchbury, 1993), PPT

Improving Scientific Presentations: 10 Commandments (Eray Aydil, 2025)

Posters

Ten simple rules for a good poster presentation (Thomas C. Erren, Philip E. Bourne, 2007)

Optimize your conference poster's impact (AIChE 2010)

Speaking Tips for Presenting a Poster (ACS, 2009)

Tips for Presenting an Undergraduate Poster (ACS, 2012)

Advice on designing scientific posters (Colin B. Purrington, 2006)

Poster Presentations (Hilary M Jones)

Mortal sins in poster presentations or, How to give the poster no one remembers (Thomas G. Wolcott, 1997)

The DOs and DONTs of poster presentation (Steven M. Block, 1996)

Style

Citations: Use a simplified style: “Last Name 1, Last Name 2, …, J. Abbr., Year, Vol, First Page”. When the author list is too long (e.g., >3), use “Last Name 1, et al.”

Preparing Manuscripts

author — Tue, 06 Dec 2016 00:04:10 GMT

Getting Started

Any writing or presentation should reflect our own thought process on why we initiate the project, how we design the research plan, how we interpret the results, and how we troubleshoot where needed. So they should probably start by describing what problems we are trying to understand; What people already knew / what the state of the art is; What calculations we did and how you did them; What you found; What the findings tell you.
Learn from others: writing and presentations.
See Fundamentals of Data Visualization for a nice introduction to data visualization and techniques for visualizing different types of information.

Checklist

Correct use of hyphens, em dashes, and en dashes
When formatting equations and symbols, only italicize variables (but not parentheses/square brackets, operators, or labels) italic; e.g., "E / k_B [K]" should have "E" and "k" italic, but not the division sign, label "B", or the square brackets.
For journals with a double-column layout, display items such as figures and tables should target either full width (~6.5 -- 7 in) or column width (~3.25 -- 3.5 in). This will signal to copy editors that we prefer them to occupy a single column or across two columns.
To manage the placement of display items, place them and their captions into a text box with Shape Format -> Shape Outline and Wrap Text set to "None" and "Square", respectively. Preferably, the text box is then positioned to the top or the bottom of a page by using the Shape Format -> Align option. Sometimes, Word will not break a continuous paragraph to use up the whitespace in the previous page. To fix this, move the text box below the paragraph, which will allow the paragraph to flow to the previous page, and then re-align the text box to the top of the page.
Journal names are properly abbreviated. Authoritative sources for journal name abbreviations include ISSN List of Title Word Abbreviations, Web of Science Journal Title Abbreviations, and Abbreviations for Elsevier journals.
Double check the references. References imported to Zotero by DOIs often have extra whitespace in front of subscripts (e.g., "CO ₂").
Follow journal guidelines and if available, use journal-specific Word/Zotero/LaTeX/BibTeX templates. For initial submission, when citing an article with 10 or more authors, list the first 10 authors' names and then use "et al." to represent additional authors (ACS Style Guide).
When you address a comment in Word, Overleaf, etc., please take care not to remove it. If you delete and rewrite the block of text on which a comment is made, then the comment itself will also be deleted, making it very hard to follow up on them. One way we can avoid this is by first writing new text inside the commented block and then delete the old text. Another way for Word documents is to use strike-through text for the word or sentence the comment is associated with and write/edit around the strike-through text.

MS Word

Do NOT make stylistic text changes on individual blobs of text, including title and headings. For example, do not use these ribbon options to make heading text bold and centered. This is a bad practice because you have to 1) make the same style changes for all headings; 2) remember to do the same for any new sections that are added later; and 3) edit every heading all over again if you ever need to submit to another journal with different requirements on formatting.
USE appropriate Styles to mark semantics of the text (e.g., title, Heading 2, Heading 3). If say, Heading 3 does not have the desired format, click Modify to edit in the dialog that pops up. You can make more changes by clicking the Format button. A nice side effect of this is that you can turn on Navigation pane and see a clear outline of the document.
NEVER hard code Figure X / Table Y in text. Use Cross-reference. To customize cross-reference labels (e.g., "Fig." or "FIGURE" instead of "Figure"), create a new label. Do NOT make changes to the labels manually, which will not work with Cross-reference properly.
See here for how to combine multiple references. That is, do not write "Figure 1, Figure 2, Figure 3"; instead write "Figures 1 -- 3" or "Figures 1, 2, and 4" (if not continuous).
To add automatic equation numbering, use either #() syntax (available since Word 2016) or 3-column no-border tables with captions. To cross reference equations, first select the equation number and insert a bookmark. You can then use Cross-reference.
To cross reference section numbers (e.g., Section 2), in the Cross-reference dialog, select Heading as Reference type and Heading number as Insert reference to. For this to work though, the numbers in section heading (e.g., "2. Methods") must be added using the numbered list functionality, rather than by hand.
The above points are very useful when adding new figures as the document will auto-update the figure/table numbering. To update cross-reference labels (in the text itself, not the captions), right-click the label and select "update field" for manual updating. See here for updating all fields at once.
NEVER use text boxes for figure labels. Text boxes often move out of alignment with the figures they are supposed to annotate. Instead, annotate the figure in PowerPoint. When you are done, select everything for the figure, right click, and either 1) Save as Picture or 2) Copy, right click in Word, and Paste as Picture.

TeX/LaTeX

Special symbols

Text mode Angstrom: \AA Math mode: Angstrom: \mathring{A}, \r{A}, \r{\rm A} degree symbol: ^{\circ}

For quick referencing, Kire Labs made a tool called Dexterify which uses symbol recognition to identify which LaTeX commands will give the user-drawn character. For example, you can draw a tilde symbol in the drawing box, and Dexterify will output various options for tildes.

Figures

When embedding figures in latex, the float default is usually h for "here." For publications this should be t for "top." For figures that are too large, the default behavior is to put the figure at the end. Putting an exclamation point will gurantee this:

\begin{figure}[t!]

Controlling figure and table placement in LaTeX explains how to place float objects to places they belong.

For the Xmgrace files, use figures in .eps format. Make sure that the figure has the correct size in Xmgrace by providing the correct number of pixels. Font size in labels and legends should be commensurate with text size in the journal.
ACS journal column widths are 3.25 in. So make sure to never go below that in width. It is not necessary to make the figures 3.25" in size unless otherwise requested by the journal.

Software

When preparing manuscripts, keep all LaTeX as simple as possible. Do NOT use any packages and only use the documentclass provided by the journal. Always follow the author guidelines from the journal as closely as possible.
Neither ACS, AIP, Elsevier, nor Nature use "pdflatex". Thus, we must use regular "latex".
Data management plans now require extensive Supporting Information. Essentially all simulation data that appears somewhere in a figure, also need to be as numerical data in a table in the supporting info. This is quite labor intensive.
Relative paths should be used
Some publisher programs will have difficulty with lines of longer than 255 characters.
Make sure to name the PDF something appropriate. For example, the ACS supporting info file name is named based on the main paper. So, for example, don't submit full_article.pdf
ACS journals have a mistake with mathematical symbols. The statement

The result is about 5.2

would be translated to

The result is ~5.2

This is the wrong symbol for about; you should use the approximate equals yourself; the program will think it's a mathematical equation instead and it will correctly translate.

Latex differentiates between opening and closing quotes.

BibTeX

Correct format

An example BibTeX entry that has multiple issues:

@article{X22,
author = {Gupta, Prashant Kumar and Meuwly, Markus},
journal = {The journal of physical chemistry. B},
number = {35},
pages = {10951--9},
volume = {116},
year = {2012}
}

Title: This is sometimes missing from the output of reference manager software, but is needed for all proposals and for some journals (e.g., ACS journals now want the titles for citations). Also be careful that the Web of Science may use the wrong capitalization: only the first word and names should be capitalized (sentence case). The BibTeX style file should change titles to the journal format automatically (e.g., title case for ACS journals). However, there is a problem with names. For example, the name of our method has to be typed as "{M}onte {C}arlo" otherwise the first letter of these names will be lowercase.
Author: Edit so that only the initials are included. No journal wants the full middle name of the authors. Also NEVER use "et al." in your .bib file. For the few journals that want fewer names, we will modify the BibTeX style file or directly edit the .bbl file.
Journal: Edit to use the correct abbreviation. LaTex does not automatically find the correct abbreviations for the journal title. When using Zotero, double check "Journal Abbr" is set correctly.
Number: Remove or make sure the BibTeX template removes them. There is no single journal that requires the issue number in the list of references.
Pages: Page numbers need to be given in full. That is, "10951-9" should read "10951-10959".
Some journals do not allow a separate .bib file, in which case the references need to be embedded in the main .tex file.
When a problem is found for a reference, make sure to fix the citation source in Zotero or BibTeX. Develop the habit of checking the title and journal abbr fields whenever you import a new reference, either by DOI lookups or by extracting PDF metadata.

Writing

Skillful writing of an awful research paper (Royce Murray, 2011)

Paper writing gone Hollywood (2017)

Siepmann's comments on clarity and precision of language

Visualization

author — Tue, 19 Feb 2013 00:39:59 GMT

Overview

See Fundamentals of Data Visualization for a nice introduction to data visualization and techniques for visualizing different types of information.
Molecular visualization
- Jmol: an open-source Java viewer for chemical structures in 3D with features for chemicals, crystals, materials and biomolecules. Useful for quick inspection of structure and trajectories.
- OVITO: a scientific data visualization and analysis software designed for molecular and particle-based simulation models.
- VESTA: a 3D visualization program for structural models, volumetric data such as electron/nuclear densities, and crystal morphologies. Convenient to use with VASP.
- Multiwfn: a very powerful program for electronic wavefunction analysis (e.g., charge density, ELF, ESP, NCI, and orbital fields).
- VMD: a molecular visualization program for displaying, animating, and analyzing large biomolecular systems using 3-D graphics and built-in scripting.
- PyMol: a molecular visualization system for rendering and animating 3D structures.
Structure editor
- Avogadro: a free and open source molecular editor and visualization tool.
- Materials Studio
- Schrodinger
- Aten: a molecular modelling package that allows editing and manipulation of atoms and molecules in isolated and periodic systems, and offers full editing via either graphical user interface or command line (scripted) operation.
3D visualization
- ParaView and VisIt are two general purpose visualization tools, which support volume rendering and other advanced techniques.
- VTK is an open-source C++ library for 3D computer graphics, image processing and visualization, and has several interpreted interface layers including Tcl/Tk, Java, and Python.
2D plotting

Materials Studio

Ask a group member if you need to use this.
From the Help -> Help Topics menu, you can find a list of help topics and tutorials. I recommend going through "Tutorials > Visualizer tutorials" and "Materials Studio interface > Mouse and keyboard actions".
Browse "Sketching and editing structures", "Visualizer Tools", and "Building Tools". Familiarize with its capability and come back to these topics when you use them.

Count atoms in a unit cell

Right Click -> Display Style -> Lattice Tab -> Style: Original
Edit Menu -> Atom Selection -> Select by Property: Element, Selection Mode: Create a new selection -> Select

Schrodinger

Obtain the required software license files through your institution or the software vendor before installation. Follow their licensing terms and keep the license files private.
Download and install Schrodinger.
After installation is complete, start "Configure Software" and add the two .lic files extracted from Schrodinger_licenses.zip.
You can now launch Materials Science (use this instead of Maestro).

NOTE: For licensing to work in the last two steps, your IP needs to belong to the University network, so use VPN if you're not on campus. After the program is started, you can disconnect from VPN.

From the Help -> Tutorial menu, you can find a list of tutorials and click View to see it in browser. I would recommend going through "Introduction to Maestro for Materials Science".

VMD

Installing VMD

Go to the website. Click on the VMD button and find your platform (preferably with Open GL). Click to begin the download.
Follow the instructions to create a user account (used to determine usage and reported to the NSF or other funding agency supporting VMD development) and browse the README file (onscreen) while the program downloads.
When download is complete, open a Finder window and move VMD into the Applications folder. To do this, drag the VMD icon (looks like a water molecule) over the ‘Applications’ menu on the left of the Finder window. The cursor should become a green plus sign.

Generating high-quality images

Make sure that in the Graphical Representations window, all your visualized components have high values for their resolution.
Open File Render Controls: from VMD Main window, File -> Render.
Changing the method from "Snapshot (VMD OpenGL window)" to basically anything else will dramatically improve the image quality. VMD User's Guide recommends "Tachyon (internal, in-memory rendering)". The default Snapshot renderer just captures what is on the VMD OpenGL Display area. It will even include other objects that are on top of the display window at the time when you perform the rendering.
To use renderers other than "Snapshot (VMD OpenGL Window)" or "Tachyon (internal, in-memory rendering)", you need to perform two steps: first generate a scene description file from VMD, and second run the separate rendering program with your desired options to generate a raster graphics file. See Chapter 8.2 Higher Quality Rendering of VMD User's Guide for more details on the different rendering programs and recommended options.
For example, to use Tachyon renderer with custom parameters, choose "Tachyon", and in the Render Command text box enter: "/Applications/VMD 1.9.1.app/Contents/vmd/tachyon_MACOSXX86" -aasamples 8 -rescale_lights 0.4 -add_skylight 0.7 %s -format TARGA -o %s.tga. VMD will generate a scene file suitable as input for the Tachyon rendering program, with the name specified in the Filename text box, and then run the command given in the Render Command box to actually render the image. "%s" is the file name specified in the Filename box.
In addition, you can manually modify the scene file in order to change settings such as the resolution, anti aliasing threshold, and re-run the rendering program from the command line to generate another image file with the modifications you just made to scene descriptions.

Add/remove bonds

When reading in a simulation snapshot, VMD will try to determine the bonds between atoms. Sometimes it will draw bonds where they should not exist and at other times it can also miss certain bonds. Use the pickbond script from VMD Script Library to add/remove bonds.

Save the file to your disk.
Open VMD Main Window -> Extensions -> Tk Console.
Change to the directory where you save the tcl script and invoke the script:

cd target/directory
source pickbond.tcl # read in the script
pickbond add/del/stop # add bond, remove bond, stop pickbond action

Change the mouse mode to Pick by pressing P or from VMD Main Window -> Mouse -> Pick.
Click the two atoms between which you want to create and delete a bond.

Perform periodic boundary conditions at arbitrary positions

Sometimes you will want a particular molecule to be at the center of your system while your movie file has it always at the boundary of the simulation box. Suppose the simulation box is cubic with sides of length 32.0, and the molecule of interest is at (16.0, 32.0, 0.0), you can do the following:

set sel [atomselect top "all"] # select all atoms in your top molecule
$sel moveby {0 -16 16} # move all molecules
pbc set {32.0 32.0 32.0 90.0 90.0 90.0} # set the lattice parameters of the simulation box
pbc wrap # perform periodic boundary conditions
pbc join connected # re-connect molecules

ChemDraw

Start by selecting an appropriate template: File -> Apply Document Settings from -> ACS Document 1996.
Arrange drawings properly with respect to the page/column width or whatever is the desired final figure width. Do not scale pasted image in Word or LaTeX; otherwise the font and symbols will not be of the right size.
Select new objects you draw and try auto-adjusting bond lengths and angles: Structure -> Clean Up Structure.
To add labels, right select the structure and click Analysis -> Structure, which will create an auto-aligned label below the structure. You can then edit the label. This will stop auto-updating. Labels created this way will not have warnings about wrong valency, etc. Alternatively, right select labels created using Text tool and uncheck Interpret Chemically.
Make sure the schematic makes sense. For example, if you have a two-step reaction: A + B <-> C and then C + D -> E, do not draw A + B <-> C + D -> E. Instead, introduce D using a half arc above the second arrow.
See this video on how to quickly draw a catalytic cycle and a few handy tips.

Plotting Data

author — Mon, 19 Sep 2011 08:34:59 GMT

Overview

See Fundamentals of Data Visualization for a nice introduction to data visualization and techniques for visualizing different types of information.

Color selection

Simple tools for mastering color in scientific figures

Choosing color palettes for scientific figures

Color In Scientific Figures (AJE, Stacie Meaux)

Coloring for Colorblindness

Color palette

Color palettes for colorblindness

GeoDataViz Colour Palettes

Adobe's color wheel

ColorBrewer: Color Advice for Maps

Color palettes with hex codes (Matlab friendly)

Xmgrace

Creating the default settings for new graphs

Open xmgrace, make the desired changes to the various settings, save the file as ~/.grace/templates/Default.agr. A template appropriate for publication can be downloaded here. Note that in order to change the actual page dimensions, you need to go to View -> Page setup, and in the Page section, change Size to "Custom" and the Dimensions unit to be "in", put 3.25 x 3.25 (requirement for half-page figure) or 7 x 7 (for full-page figure). This will fix the size on the output devices (X11, PDF, PNG, or actual printers). The resolution on the computer screen (X11) is set in the Resolution (dpi) text box, and the resolutions for other types of devices are set in File -> Print setup -> Resolution (dpi). Unfortunately, the resolution setting and the default print device are not saved in the template file, but see below.

Setting the default printer to print to .png files with 300dpi

Create the file ~/.grace/gracerc.user and enter the following text:

HARDCOPY DEVICE "PNG"
DEVICE "X11" DPI 192
DEVICE "PNG" DPI 300
DEVICE "PDF" FONT ANTIALIASING on
DEVICE "EPS" FONT ANTIALIASING on
DEVICE "PostScript" FONT ANTIALIASING on

Special symbols

Italic, bold

italic: \1italic
bold: \2bold
italic and bold: \3xyz

Subscript, superscript

subscript: 3\s10\N
x-squared: x\S2\N

\N returns the font position to normal.

Special symbols

Angstrom symbol: \cE\C

For other characters, look at this list: ascii table with low and high characters. Just use the character from the left column between \c and \C to produce the one from the right column. I highlighted the most interesting characters (for a scientist).

Greek letters

theta: \xq\0
rho: \xr\0

\0 returns the font type to normal.

The \c, \C, and \x options are listed as deprecated in the xmgrace manual. The new method to insert special characters in xmgrace is:

Press ctrl-e while positioned in a text-edit field to bring up the font dialog box.
Select the desired font from the drop-down list. You probably want to use Symbol because it contains many of the commonly used special characters.
Click on the character you want to insert

This will put something like

\f{Symbol}q\f{}

into the text box.

Gnuplot

Creating the default settings for new graphs

You can use a template file to get a good default and ready-to-use color palettes for new graphs by putting the following line at the top of a new .gpl script.

load 'path/to/default.gpl'

Ternary diagram

You basically need to calculate the (x,y) coordinates for all your data points as well as the three vertices of the triangle. This template file will produce something like this

Notes on figures

When in doubt, check a recent publication in the target journal.
Remove graph title.
For a figure with multiple panels, assemble them in PowerPoint and export it as a single image of the correct dimensions.
For a two-column layout, a single-column figure should have a width of ~3.25 in and a full-page figure should have a width of ~6.5 -- 7 in, with at least 300 DPI. Make the image itself the correct size, instead of resizing it in Word or LaTeX. Some journals require figures to be uploaded separately and the sizing information in Word may be ignored.
When displayed at the correct size, check that the fonts in the figure are not too small. They should be the same as or one point below the font size of the main text.
Use the correct negative sign (in Xmgrace use Symbol font for label marks).
Format variables (but not parentheses/square brackets, operators, or labels) italic; e.g., "E / k_B [K]" should have "E" and "k" italic, but not the division sign, label "B", or the square brackets.
Use a consistent number of significant digits in axis labels (except 0); e.g., 0, 0.5, 1.0, 1.5, ...
Start and end the axes at major or minor tick marks.
As a general guideline, select an appropriate increment so that there are between 3 and 6 tick marks on each axis.