Hythem Sidky
@hythemsidkyPostdoc @UChicagoIME | Liquid crystals, molecules, proteins, data, probability, physics, and ML... I roll (digital) dice for a living. I also climb rocks.
Similar User
@CellardEleonore
@iqsaweb
@emrane
@a_el_shamsy
@IslamicOrigins
@YetAnthrStudent
@Rurouni_Phoenix
@PhDniX
@RamonIHarvey
@MohsenGT
@shahanSean
@m_arabia
@ShAmmarKhatib1
@TheIsmaili
@shakerr_ahmed
(1 of 3) Happy to share version 1.0 of our SGOOP code for finding improved reaction coordinates for enhanced sampling of rare event systems starting from preliminary, approximate metadynamics runs using trial coordinates, or even long unbiased MD runs. github.com/zwsmith200/SGO…
Amazingly common at scientific meetings these days: response to introducing two people, “No we haven’t met, but we know each other from twitter.”
I'm organizing a session with @JuandePablo6 on Free Energy and Reactive Path Methods for the 2019 @APSphysics March Meeting. The deadline is next Friday, earlier than usual. Please retweet if your followers might be interested, and contribute if you are. goo.gl/EkT7Th
Extremely proud of my graduate student Vikramjit Rathee for his latest paper, just accepted to @J_A_C_S, on the complexation of weak polyelectrolytes. Ben Sikora and @hythemsidky with big contributions as well. pubs.acs.org/doi/10.1021/ja…
LaTeX is wild
Fabulous talk on #catalysis and the nitrogen cycle in @WPI Chem Eng by Prof @wmfschneider of @NDCBE. Bill was a PhD student with me and I am so impressed with and proud of all he has accomplished! #ProudPapa
I wrote a script to grab 2 months worth of car insurance quotes from @Ask123ie to show how they adjust your premium based on the day of the week you were born. Here's the graph, see if you can spot the pattern Spoiler: It costs €130 more if you were born on a Sunday vs a Monday
Quick phishing demo. Would you fall for something like this?
This is always a potential issue with open-access publishing, though it is also an issue with the proliferation of journals in general, as content becomes ever more thinly spread. sciencemag.org/news/2018/09/o…
In 1923, 2 scientist at University of Toronto (U of T) got the Nobel Prize for discovering insulin. The patent was sold to U of T` for $1 so insulin would be widely available. Today, a single vial of insulin is > $250 and most patients use 2-4 vials per month and can't afford it
The soaring cost of insulin, which is killing diabetics, should be a front-page scandal. #insulin4all npr.org/sections/healt…
Never confuse kindness as a sign of weakness. twitter.com/lad/status/888…
Possibly the biggest / most important change is the addition of ColumnTransformer to handle heterogeneous pandas dataframes. Right now you need to provide a callable (or list) to say what columns are, say, categorical. Next up: providing robust mechanisms in sklearn.
Help us by testing the scikit-learn 0.20 release candidate! scikit-learn.org/0.20/ Install via "conda install scikit-learn=0.20rc1 -c conda-forge/label/rc -c conda-forge" from conda-forge or with "pip install --pre scikit-learn " from pypi!
There is a difference between hard work and emotional abuse. Grad school should be challenging & intellectually difficult. I'm not arguing it should be easy. You should work hard, but you should NOT be emotionally abused. The culture should not promote emotional abuse.
Help us by testing the scikit-learn 0.20 release candidate! scikit-learn.org/0.20/ Install via "conda install scikit-learn=0.20rc1 -c conda-forge/label/rc -c conda-forge" from conda-forge or with "pip install --pre scikit-learn " from pypi!
Our September issue is live! Modelling kinases (on the cover), microporous membranes, strained azacyclic allenes, designer enzymes, proton hydration structures (and a great cover line related to them!), N-doped graphdiyne electrocatalysts & much more... nature.com/nchem/volumes/…
Released v2.1 of my tensor contraction order optimizer, opt_einsum! Added @PyTorch backend support and @TensorFlow eager-mode execution along with other optimizations. Downloads available from #pypi and @condaforge optimized-einsum.readthedocs.io/en/latest/ #compchem #numpy
Apologies. I generally find deep-learning practitioners holding court as public intellectuals on twitter to be a bit tiresome.
Here's a thread on basic tips to help improve your next science figure! 📝 (ie common mistakes I see). It'll cover: 1) Contrast (color value) ⬛️⬜️ 2) Color (it’s a tool, not decoration!) 🌈 3) Fonts 🔤 4) Image Resolution 📷 5) Spacing / margins 📐 6) By request! 💡
Improving Chemical Autoencoder Latent Space and Molecular De novo Generation Diversity with Heteroencoders arxiv.org/abs/1806.09300 #compchem
Liquid crystal display display 😬
United States Trends
- 1. Hunter 2,17 Mn posts
- 2. Cyber Monday 73,3 B posts
- 3. Take Ctrl 1.603 posts
- 4. #GalaxyFold6 N/A
- 5. $CUTO 11,3 B posts
- 6. #GivingTuesday 5.380 posts
- 7. #IDontWantToOverreactBUT 1.743 posts
- 8. Enron 11,2 B posts
- 9. RTFKT 5.945 posts
- 10. Kevin Warren 2.089 posts
- 11. AP Poll 7.264 posts
- 12. Burisma 81,8 B posts
- 13. Thomas Brown 3.098 posts
- 14. Miller Moss 1.114 posts
- 15. #Duolingo365 16,5 B posts
- 16. Tony White 1.649 posts
- 17. Intel 30,9 B posts
- 18. #MondayMotivation 19,7 B posts
- 19. Daniel Penny 23,4 B posts
- 20. Bloodborne 6.916 posts
Who to follow
-
Dr. Éléonore Cellard
@CellardEleonore -
IQSA WEB
@iqsaweb -
Emran El-Badawi
@emrane -
Ahmed El Shamsy
@a_el_shamsy -
Joshua Little @islamicorigins.bsky.social
@IslamicOrigins -
Yet Another Student
@YetAnthrStudent -
Rurouni_Phoenix
@Rurouni_Phoenix -
Marijn van Putten
@PhDniX -
Ramon Harvey
@RamonIHarvey -
Mohsen Goudarzi
@MohsenGT -
Sean W. Anthony
@shahanSean -
مجْمع اللغة العربية
@m_arabia -
Ammar Khatib, PhD
@ShAmmarKhatib1 -
The Ismaili
@TheIsmaili -
Ahmed Shaker
@shakerr_ahmed
Something went wrong.
Something went wrong.