Dhanshree Arora | 🌈
@dhanshree_arora
Followers
192
Following
1K
Media
8
Statuses
137
Your friendly neighborhood programmer rambling about the new cool thing I did | She/Her https://t.co/v05uODlkgb
India
Joined March 2015
My favorite issue on the internet right now: requests gets updated, breaks docker_py, and many projects go into utter disarray overnight. Brb, going to go pin our dependencies now.
github.com
With requests 2.32.0 (released about an hour ago as I write this), the docker library as called by tox-docker fails with the following exception: Traceback (most recent call last): File "/home...
0
0
2
I love how the govt reminds me everyday to file my taxes. One month of those emails finally got me to schedule it on my calendar. Thanks @IncomeTaxIndia that's the kind of babysitting one needs sometimes.
0
0
7
This led us to computing entropy for different functions (the hypotheses functions) and it was pretty apparent that the exponential distribution of ISI contained the most amount of information by comparison. [lil secret: I need to revisit this bit again] (6/n).
0
0
1
Finally, with the why model, we looked at why the exponential distribution is the more optimal choice, and why not, eg, a linear function? The idea is to convey as much information as possible (ie highest entropy) (5/n).
1
0
0
When working through the how model, we modeled the underlying mechanisms as closely as possible (presynaptic excitations, inhibitions, as well as a leakage factor within the membrane potential of a neuron) to achieve the function we observed within the what stage. (4/n).
1
0
1
We looked at inter-spike-interval (ISI) data from Steinmetz et al. (2019) Within the what model, we plotted histograms of this data and observed that the data shows an exponential distribution (3/n).
nature.com
Nature - Recordings from 30,000 neurons in 42 brain regions are used to delineate the spatial distribution of neuronal activity underlying vision, choice, action and behavioural engagement in mice.
1
0
1
How models: What are the underlying mechanisms that lead the function observed in the what model. And finally, the Why models: why did evolution decide this was the most optimal way to do things? (2/n).
1
0
2
Chronicling my CNS journey with @neuromatch academy. Yesterday (W1D1), we learned at three important modeling approaches (what, how, and why).What models: you have a dataset and you wanna find out *what's* happening, ie which function would best describe/fit this data? (1/n).
1
0
7
Organizing events is a lot of work, but amazing people and the swags make it worth every bit.
All set for #Outreachy1000InternsCelebration #Outreachy1000InternsCelebrationIndia. @VaniChitkara @kavya_jaiswal_
11
2
90
Come work with us at @ersiliaio during the May-August round of @outreachy internships, you'll get to brush up on Python, machine learning, and a lot of bioinformatics! Oh and did I mention - you gain Open Source creds!.
1
3
9
RT @rasbt: Convolutional networks strike back, again. The fully convolutional ConvNeXt v2 extends the successful ConvNeXt architecture by a….
0
337
0
I'm gonna leave here a real fun Python anti-pattern I came across in a codebase today:. `for flag_variable in [True]:`.
2
0
6
Sage advice, really. Document for thyself first!.
0
0
2
TIL: SMILES to represent chemical substances in ASCII notation. There are a whole bunch of algorithms dealing with this bc you've to count for isomers, canonical forms (recall Benzene rings) & other cool stuff. There's also a whole intersection w/ graph theory & CFL! @ersiliaio.
1
0
0