import pandas as pd
import matplotlib.pyplot as plt
import matplotlib.dates as mdates
import seaborn as sn
df = (pd.read_csv("/tmp/overview_2022-03-21.csv") # i just used curl beforehand
.assign(date=lambda x: pd.to_datetime(x["date"]))
.set_index("date")
.melt(value_vars=[
"newCasesBySpecimenDate",
"covidOccupiedMVBeds",
"newAdmissions",
"newDeaths28DaysByDeathDate"],
var_name="Data", ignore_index=False)
.assign(Data=lambda x: x["Data"].replace({
"newCasesBySpecimenDate": "New Cases",
"newAdmissions": "Admissions",
"newDeaths28DaysByDeathDate": "Deaths",
"covidOccupiedMVBeds": "Ventilated"
}))
)
ax = sn.scatterplot(data=df, x=df.index, y=df["value"], hue="Data")
ax.set(xlabel="Date", ylabel="Daily rate", yscale="log")
ax.xaxis.set_major_formatter(mdates.DateFormatter("%b"))
plt.show()
I spend 2 minutes on the pandas part and 20 minutes on the plotting part, which really says it all. Seaborn's support for smoothing is really bad and doesn't play nicely with datetimes for some reason, so if I wanted smoothing I'd need to do it myself. And the other stuff I left out requires going into matplotlib's documentation which I don't want to spend time on.
pandas is as good or better than R's dataframe manipulation, but R's plotting tools are best in class. I hate all the python plotting libraries.
o https://github.com/asavinov/lambdo - Feature engineering and machine learning: together at last! The workflow engine allows for integrating feature training and data wrangling tasks with conventional ML
Could you recommend modem / router equivalents ( whichever applies to replacing ISP supplied equivalent, I'm assuming both? ) at a few different price brackets?
Ubiquiti gear appears to be quite dear, alternatively are they worth the expense? Are there any benchmarks etc which demonstrate this, or further reading?
There's the tech question and the choice question.
Tech: I use an open-source knowledge graph system, Semantic Synchrony:
https://github.com/synchrony/smsn/wiki
It uses Neo4j, so it's fast and it scales enormously; it offers a true graph, not a tree; it offers full-text boolean search via Apache Lucene; it lets you connect public and private notes in the same graph, and lets you merge (some or all of) your graph with someone else's.
Choice: I believe in introspection and sharing, and not willpower. If I feel like I have to force myself to do something, it is because I am at war with myself. Instead I meditate, journal (using SmSn), do "knowledge gardening" of the earlier things I've journaled (using SmSn), try to talk to different kinds of friends, for different kinds of perspectives.
I studied economics, and it changed my life, and I recommend it to everybody. It's hard to make good decisions in the labor market, for example, without a deep understanding of it. But econ is not just about money; it's about constrained optimization generally, and applies to time management, reading, travel, love, literally everything we do.
Same here. I bought a Philips BDM4065UC monitor in 2015 (40 inch, 3840x2160 resolution) for 740 € (in Germany) and I'm loving it. The only cave-at with that model is that reaction time is a bit slow (I would guess 5 ms between full black and full white).
When I replace it, the replacement will be the same size and resolution, but I'll probably go for OLED and HDR-10 or better.
I've tried a lot of these tools. The trouble is that things go extremely well until you hit a wall. With no way to extend them with something like freeform code, the wall doesn't move.
The best ones, then, tend to be the ones that have been around long enough, or had enough foresight to have covered various corner cases.
The other issue is that most of them have "per user / per month" pricing, which doesn't work for many use cases. Things like, for example, an applicant tracking app. If you use the functionality to a high degree, every job applicant is another end user license.
Sell something. Anything. Not just software or online. Just like in code, it takes 10,000 hours of practice. If you need help getting started shoot me an email. No bs or strings.
"The Power of Now" by Eckhart Tolle and "Starting Strength" by Mark Rippetoe (applicable to men mostly) contain lifetime lessons on mind and body management. "Principles" is very useful for mental models and just critical thinking (or, rather, structured self-doubting).
For those who've built tech companies up from the 10 to 1000 people range, there's a lot in that link that's very easy to recognize.
If you interpret back from the more traditional business lingo, you will recognize key 'iterative development' ideas applied outside engineering.
This allows a sizable 30 year old enterprise to handle new ideas much more as a tech startup would.
On the main article topic, instead of the article's quote, “like trying to make Ray’s brain into a computer”, I'd say as an engineer imagine if you could "run a company under a debugger." Frame it that way, and I think you could imagine some neat possibilities.
If you're very good at software development / distributed systems engineering, and think self-driving management or a self-driving fund might be even more interesting than yet another self-driving car, we're always hiring. Hit me up via profile.
That's a brilliant idea. I wish I had that article years ago. I'd even consider writing it but I've never blogged and feel it would just be a murky mess. I'd happily collaborate though. =)
pandas is as good or better than R's dataframe manipulation, but R's plotting tools are best in class. I hate all the python plotting libraries.