HNHacker News
TopNewBestAskShowJobs

ggrothendieck

94 karma · joined October 10, 2013

submissionscomments
ggrothendieck··on First-year doctors will be allowed to work 24-hour shifts starting in July
There is also the iCompare trial which is underway and to be completed in 2019.
ggrothendieck··on US government to withdraw longstanding warnings about cholesterol (2015)
The ACALM study found an association between high cholesterol and sharply reduced risk of death from 4 cancers. The numbers were large -- 22%, 43%, 47% and 30% lower risk of death. This may be due to wide-spread use of statins among those with high cholesterol. https://www.sciencedaily.com/releases/2016/07/160708123619.h...
ggrothendieck··on A Little Known SQL Feature: Use Logical Windowing to Aggregate Sliding Ranges
Actually the sqldf R package supports 4 backend databases (sqlite, h2, postgresql, mysql), not just sqlite, and postgresql does support window functions.
ggrothendieck··on Warner Bros Settles FTC Charges It Failed to Disclose It Paid Online Influencers
The HN title misquotes the actual FTC site's title. The FTC title says it "failed to adquately disclose", not that it "failed to disclose". Omitting the word adequately changes the meaning substantially. In fact, WB did disclose the payment but it was the manner in which they disclosed it that the FTC objected to -- not the failure to disclose it.
ggrothendieck··on What can doctors learn from pilots and cyclists?
The author of the Checklist Manifesto also wrote an article about this in the New Yorker: http://www.newyorker.com/magazine/2007/12/10/the-checklist
ggrothendieck··on Sweet Spot: New Study Shows Optimal Group Size for Baboons
The abstract of this paper provides a good summary: http://www.ncbi.nlm.nih.gov/pubmed/26504236
ggrothendieck··on Scientists may have solved a mystery about sea-level rise
> Is there a similar site that does the same for climate change?

It's not in Q&A form but NASA's site summarizes what is known: http://climate.nasa.gov/evidence/

ggrothendieck··on Common Probability Distributions: The Data Scientist’s Crib Sheet
The image beside each node is nice but the meaning of the edges needs to be well defined for this chart to be useful. An interactive graphic of 76 probability distributions which does define the edges and also links each node to a description is available here: http://www.math.wm.edu/~leemis/chart/UDR/UDR.html and is further discussed here: http://www.amstat.org/publications/jse/v20n3/leemis.pdf

Wikipedia also has a chart: https://en.wikipedia.org/wiki/Relationships_among_probabilit...

ggrothendieck··on After Capitalism?
The following link is another review of the same book by Paul Mason http://www.paecon.net/PAEReview/issue73/Gillies73.pdf but it focuses more on the economic idea, apparently discussed in his book, that the digital economy features a zero marginal cost of production. Now, if the purpose of the economy is to allocate scarce goods what happens when digital goods dominate the economy and have zero marginal cost (wikipedia, arxiv, github, free software, etc.)? The idea that scarcity might disappear would seem to undermine the very foundation driving economic transactions.
ggrothendieck··on Decline in the labor share of income in the U.S.
The link is set to start at 3:22 but its the whole Humans video if you start from the beginning.
ggrothendieck··on Decline in the labor share of income in the U.S.
This video discusses horse unemployment and more: https://www.youtube.com/watch?v=7Pq-S557XQU&t=3m32s
ggrothendieck··on Comparison – R vs. Python: head to head data analysis
For R: (1) instead of `sapply(nba, mean, na.rm = TRUE)` use `colMeans(nba, na.rm = TRUE)`. (2) instead of `nba[, c("ast", "fg", "trb")]` use `nba[c("ast", "fg", "trb")]`, (3) instead of `sum(is.na(col)) == 0` use `!anyNA(col)`, (4) instead of `sample(1:nrow(nba), trainRowCount)` use `sample(nrow(nba), trainRowCount)` and (5) instead of tons of code use `library(XML); readHTMLTable(url, stringsAsFactors = FALSE)`
ggrothendieck··on Ask HN: How to Find the Best Info on Cancer Treatments?
Enter the disease name in google and one of the first results will likely be the medical association that deals with the disease. Check out its web site. Others in the same situation are a good source of info so google for the disease name along with "support group" and also try the disease name and "forum". Abstracts of the medical literature summarize the state of the art and recommended protocols. Abstracts to all medical literature can be found at http://www.ncbi.nlm.nih.gov/pubmed . Enter the disease name in the search box on that page and then in the left hand panel click on Reviews to get all the abstracts of review papers. Registered clinical trials are shown at: http://www.clinicaltrials.gov . In the search box on that page enter the disease name.
ggrothendieck··on Miller is like sed, awk, cut, join, and sort for name-indexed data such as CSV
Regarding quoting support, check out https://github.com/dbro/csvquote
ggrothendieck··on Soloshot: Automatically film outdoor activities without a camera operator
Not the same but there exists a microphone which can be thought of as an audio version of this: https://www.acousticmagic.com/ It senses the direction of the speaker and outputs audio from that direction filtering out audio from other directions. As different speakers in different locations speak it changes the direction it uses. It also outputs a location signal that can be used by software to, for example, have a computer position a camera toward the current speaker. (I don't have one - this is just based on the material I have read.)
ggrothendieck··on Simple Rules for Healthy Eating
Regarding rice, Consumer Reports (CR) has been urging the FDA to set limits on rice consumpton due to the arsenic levels in rice that they have measured. A certain amount of this is due to naturally occurring arsenic but pollution from industiral production of goods could add more. In the meantime CR has issued their own point system based on the arsenic content of rice: http://www.consumerreports.org/cro/magazine/2015/01/how-much...
ggrothendieck··on Gift of the Moon, Bane of the Spanish: The Story of Yerba Mate
Note link between Yerba Mate and esophageal cancer: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3947123/ and http://preventcancer.aicr.org/site/News2?page=NewsArticle&id...
ggrothendieck··on When arts die, they turn into hobbies
Leonard Cohen managed to monetize his poetry through his music.
ggrothendieck··on Starvation Based Chemotherapy: A Novel Approach for Cancer Treatment [pdf]
There is a site and free self published e-book of about 60 pages (hard copy is available too) on an IF diet in which one fasts for all but 5 hours a day for the purpose of weight loss. The book describes the author as a retired medical doctor and former researcher in the Metabolism branch of the National Cancer Institute. Cancer is not discussed in the book although there is a brief section on physiology and the three mechanisms throught to underlie the effect. See http://www.fast-5.org/
ggrothendieck··on Bad luck, bad journalism and cancer rates
The fact that species with more cells don't get more cancer is known as Peto's paradox. Perhaps it is also true at the tissue level. See http://en.wikipedia.org/wiki/Peto%27s_paradox
ggrothendieck··on Adult-onset diabetes, obesity cured in lab mice, scientists report
Over the years that I have looked at Science Daily it has always provided sources and citations. I don't think there has been any change.
ggrothendieck··on Adult-onset diabetes, obesity cured in lab mice, scientists report
This is a bit harsh on Science Daily. In fact, Science Daily is much better than many other sites of its kind because it does consistently give its sources and provides citations as it did here.
ggrothendieck··on The R language, for programmers
To be clear this applies to data frames, not the language as a whole. If you are creating a data frame by reading in data then `read.table(...whatever..., as.is = TRUE)` will read in character columns as character strings and not as factors. `data.frame()` has a `stringsAsFactors` argument which can be set to `FALSE` and one can change the global default with: `options(stringsAsFactors = FALSE)` .
ggrothendieck··on Starting data analysis with R: Things I wish I'd been told
(!) sqldf works with H2, MySQL and PostgreSQL too - its not limited to SQLite. (2) Most alternatives to SQL available in R don't work well with complex multi-way joins. You wind up materializing the full outer join first which can be quite problematic. (3) sqldf is fast enough and R is slow enough that quite often sqldf is faster than base R so if R is fast enough for you then its likely that sqldf is too. (4) The speed is often the last thing you need to worry about. The ability to express a query in a familiar way (if you know SQL) is often the more important consideration than having to learn a new system.
ggrothendieck··on Use Python in Excel without add-ins
To run a python function you do need to write one line of VBA code and you also need a vba file that comes with the package so although its not very much to add on the VBA side the title here seems a bit misleading.
ggrothendieck··on Execute SQL against structured text like CSV or TSV
The R package sqldf together with backend packages RSQLite (the default), RH2, RMySQL and RPostgreSQL can be used in a somewhat similar manner automatically generating create statements, uploading referenced tables to a database and performing the specified SQL. Tables referenced in the SQL statement are assumed to be R data frames or external files depending on whether an R data frame or file description object of the same name exists in the R work space. The result is returned as an R data frame. See http://sqldf.googlecode.com
← PreviousPage 2 of 2