imcoconut
Joined 10 years ago
Karma
160
Submissions
108
Member since
Oct 26
Recent submissions
Comment · 1 year ago
I agree completely. Also definitions matter. Currently in the US one of the leading causes of death, if not the leasing cause, is opiods which the CDC classifies as “accidental overdose”. The healthiest people in the US are probably abusing opiods at lower rates and therefore less likely to die from the leading cause of “accidental death”. The study is from sweden but just pointing out that the assumption they make is flawed - “we should not expect a relationship between fitness and accidental death”.
Comment · 2 years ago
I thought your anecdote and commentary were relevant and extremely thought-provoking.
The person you're responding to here was clearly emotionally triggered by your anecdote. I wouldn't spend too much time trying "convince" them that what you wrote is true.
Comment · 3 years ago
it's pretty simple but on various consulting jobs I've had to build SQL databases sometimes with lot's of tables with lot's of columns. Sometimes we switch from on prem to cloud, or vice versa or switch from postgres to sql server, etc. I have this toolkit that automates a lot of the tedious stuff. it allows me to take pandas dataframes and do the following:
- auto detect and convert column types
- save as a parquet file in a folder
- then autogenerate a sqlalchemy table/metadata file in python for all tables with sensible defaults for column types (e.g. 2x the longest string in a column for varchar)
- build the db and all tables
- load data from the files into the tables
this makes it really easy to bootstrap the entire db from a folder of parquet files for testing with sqlite and then makes it easy to move to prod on postgres/sqlserver etc. Before I go to prod i still have to add constraints and keys and indexes but that doesn't take too long. and for dev/testing the data's not too big so performance doesn't really suffer from lack of keys/constraints then we can use something like alembic on the big sqlalchemy tables definition file to do db migrations.
it's kind of like this: https://github.com/agronholm/sqlacodegen but solving an inverse problem.
basically it bootstraps the db and schemas and gets me like 95% of the way there. my quality of life is better with it.
Comment · 3 years ago
This looks awesome. I'd be super interested in testing this out and providing feedback on it. I'm going to make an account and is there an email/link to submit feedback as I use it?
Comment · 4 years ago
>As far as I’m concerned Bayer and Monsanto are like evil incarnate
Especially so when they were a subsidiary of IG Farben.
>The company had ties in the 1920s to the liberal German People's Party and was accused by the Nazis of being an "international capitalist Jewish company".[8] A decade later, it was a Nazi Party donor and, after the Nazi takeover of Germany in 1933, a major government contractor, providing significant material for the German war effort. Throughout that decade it purged itself of its Jewish employees; the remainder left in 1938.[9] Described as "the most notorious German industrial concern during the Third Reich"[10] in the 1940s the company relied on slave labour from concentration camps, including 30,000 from Auschwitz,[11] and was involved in medical experiments on inmates at both Auschwitz and the Mauthausen concentration camp.[12][13] One of its subsidiaries supplied the poison gas, Zyklon B, that killed over one million people in gas chambers during the Holocaust.[b][15]
The Allies seized the company at the end of the war in 1945[a] and the US authorities put its directors on trial. Held from 1947 to 1948 as one of the subsequent Nuremberg trials, the IG Farben trial saw 23 IG Farben directors tried for war crimes and 13 convicted.[16]
Comment · 4 years ago
Not drag and drop but for a 100% python solution h2o wave is pretty cool. It has a focus on data-based dashboards but can be used to pretty easily create very good looking frontends for all sorts of applications. And you never have to leave python (whether that's a pro or a con is up for debate :))
Comment · 4 years ago
He has been saying exactly this for many years and has lead many efforts to improve both the implementation and API via projects like arrow and pyarrow.
Also, pandas was purpose built for a pretty specific domain of financial timeseries and cross-sectional data analysis at a time when the python ecosystem was much younger and very different from today.
It's not his fault it became so wildly successful! (actually it was - it's a great piece of sofware :))
Comment · 4 years ago
im a big fan of loguru.
Comment · 4 years ago
Mito looks awesome.
Just want to say that I respect the fact that you built this library, are offering it open source, for free, and want the telemetry on. You are up front and open about it.
Sometimes I think people can get a little entitled with all the work someone else puts in to a project they want to use (not accusing GP of this, speaking generally). As you said, under the license anyone is more than welcome to fork, modify and open source. Yay open source indeed :)
Comment · 4 years ago
I completely agree with you
Comment · 4 years ago
Why are you so confident?
It seems a fair number of experts have well justified doubts based on studies of already deployed lower energy radiation.
https://blogs.scientificamerican.com/observations/we-have-no...
> Citing this large body of research, more than 240 scientists who have published peer-reviewed research on the biologic and health effects of nonionizing electromagnetic fields (EMF) signed the International EMF Scientist Appeal, which calls for stronger exposure limits. The appeal makes the following assertions:
“Numerous recent scientific publications have shown that EMF affects living organisms at levels well below most international and national guidelines. Effects include increased cancer risk, cellular stress, increase in harmful free radicals, genetic damages, structural and functional changes of the reproductive system, learning and memory deficits, neurological disorders, and negative impacts on general well-being in humans. Damage goes well beyond the human race, as there is growing evidence of harmful effects to both plant and animal life.”
The scientists who signed this appeal arguably constitute the majority of experts on the effects of nonionizing radiation. They have published more than 2,000 papers and letters on EMF in professional journals.
> The World Health Organization's International Agency for Research on Cancer (IARC) classified RFR as "possibly carcinogenic to humans" in 2011. Last year, a $30 million study conducted by the U.S. National Toxicology Program (NTP) found “clear evidence” that two years of exposure to cell phone RFR increased cancer in male rats and damaged DNA in rats and mice of both sexes. The Ramazzini Institute in Italy replicated the key finding of the NTP using a different carrier frequency and much weaker exposure to cell phone radiation over the life of the rats.
> Since 5G is a new technology, there is no research on health effects, so we are “flying blind” to quote a U.S. senator. However, we have considerable evidence about the harmful effects of 2G and 3G. Little is known the effects of exposure to 4G, a 10-year-old technology, because governments have been remiss in funding this research. Meanwhile, we are seeing increases in certain types of head and neck tumors in tumor registries, which may be at least partially attributable to the proliferation of cell phone radiation. These increases are consistent with results from case-control studies of tumor risk in heavy cell phone users.
Comment · 4 years ago
This looks amazing, and very timely.
Any chance you have a mirror on gitlab (or github)?
Comment · 5 years ago
You're not necessarily wrong that the causes of the increases in prices for things like healthcare, etc. are due to policy issues, among other things. But you can kind of make that argument for any good or service - that that the increase in the price of x is a "specific industry/policy issue" and not a "monetary issue".
That doesn't negate the fact that people are still paying higher prices. If the fed is mandated with controlling the increase in prices then it is still upon them to address it, no? The fact that monetary policy is not a primary cause of the good/service specific inflation is irrelevant.
Comment · 5 years ago
This is robinhood's fault.
He's not wrong that instant (or same day) settlement would be better than T+2, but there were plenty of other brokers that did not restrict trading. This was a liquidity issue for robinhood. This is a risk you run being a "cool startup that moves fast and breaks things" in the arena of securities trading. Additionally, Some of the bugs they've experienced are absurd in the context of a broker that potentially houses people's entire liquid net worth, including a bug that reversed the direction of trades - "oh you meant to sell those shares? whoops!"[0].
Why choose robinhood when the alternatives include some of the most well capitalized institutions in the world, eg JP Morgan or Bank of America Merrill Lynch - they have literal trillions in assets each and neither has these issues. I appreciate that robinhood pioneered zero commission trades, but that has been largely adopted by the industry at this point. From my standpoint I only see risks versus their peers and precisely zero benefits.
I implore everyone here to stay far away from robinhood.
[0] https://www.nytimes.com/2021/02/02/technology/robinhood-ceo-...