Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

One in five genetics papers contains errors thanks to Excel (2016)

science.org

Post reply on HN
Print view View original

251–260 of 267 posts

Re: One in five genetics papers contains errors thanks to Excel (2016)

#251

Post by Izkata »

spockz wrote: post #116
bikenaga wrote: post #61

Earlier quoted context omitted.

No - in a subtle way, you're assuming that the unary minus has precedence, when the point is that it doesn't. You're right that "thing^2" means "thing times thing", but in "-3^2", what is it that is being squared? To write it, as you did, as "(-3) x (-3)", assumes that in "-3^2" the thing being squared is "-3". But that in turn assumes that the unary minus is done before the square. By the standard mathematical conve…

Interesting. In our high school math class when -3^2 was hand written it was ok to interpret it as (-3)^2. But maybe that could have been because spacing would be used as some informal parentheses.

It was the same for me. The question seems to be whether "-3" here represents the operation "negate" or the number "negative 3". If it's the latter, then -3^2 is (-3)^2.

Re: One in five genetics papers contains errors thanks to Excel (2016)

#252

Post by XlA5vEKsMISoIln »

tssva wrote: post #250
XlA5vEKsMISoIln wrote: post #235

Earlier quoted context omitted.

Excel chews up CSVs that it opens. I know this because an accountant checked each file our code produced using Excel before trying to import it into another program. We proofread our code before we realizing the problem was somewhere else. Shoulder-surfed the process, found the giant bug with a green X on it.

Excel doesn't change CSV files when it imports them. If the imported file was being changed then the user was saving back to the same file they imported from.

This information doesn't help anyone.

The fact is the person was double-clicking a file in a list to view its contents and Excel was trampling it. Nobody in their right mind will waste time to open Excel first, use import feature, re-navigate to the file they were already looking at, and go through the import dialog just to see what's inside.

Re: One in five genetics papers contains errors thanks to Excel (2016)

#253

Post by om8 »

Younger me would've said that WYSIWYG editors were a mistake and that researchers should've used LaTeX.

Now I think these errors are a small price to pay for convenience. One could waste a lifetime fighting small things like this and still lose. It's just the world we live in.

Re: One in five genetics papers contains errors thanks to Excel (2016)

#254

Post by wuming2 »

StableAlkyne wrote: post #110
HKH2 wrote: post #86

Earlier quoted context omitted.

Sometimes people need to be saved from themselves. E.g. spreadsheets can have mistakes very quickly when people treat them as databases and start copying columns between spreadsheets assuming that the primary keys (e.g. in one column) are identical while they happily add rows and move rows around.

To be fair, there are no good, accessible database tools around for your average non programmer user. Access tried to be this a decade ago, until MS started to let it die. So now, your only option is basically Excel. There's a reason it's the main thing people gravitate into.

Claris FileMaker was that tool. Easy to understand and provided clear separation between data and the rest. Bundled with Mac OS so everyone used it. It is now an expensive enclave.

Re: One in five genetics papers contains errors thanks to Excel (2016)

#255

Post by DiscourseFan »

RandomThoughts3 wrote: post #152
DiscourseFan wrote: post #94

Earlier quoted context omitted.

Isn't this treating 3 as something that can be analyzed from -3? Like -3 is the union of 3 and the (-), and not something onto itself. Ought it to be the case that negative is more than just a sign? Like, I think 3 should be ontologically distinct from -3, even if the addition of 3 and -3 = 0. Idk, I don't have a reason for this, it feels right for some reason.

You are mistaking the notation for the object. There is indeed two ontologically different elements 3 and (-3) in Z. The question is however purely about what is the meaning of the ambiguous without precedence rules representation -3^2. Note that it gets more complicated quickly if you want to keep thinking about it in that mathematicians often consider ontologically different but equivalent operations as the same wh…

I mean, all equations are true if all terms are set to equal zero, so the term itself must have some sort of ontological distinction.

Re: One in five genetics papers contains errors thanks to Excel (2016)

#256

Post by the_gorilla »

prepend wrote: post #177
the_gorilla wrote: post #130

Earlier quoted context omitted.

This is an awful mindset and I'm tired of people acting like tools only ever have pre-defined purpose and you need a special purpose tool to do anything. Also that's a dumb analogy. Here's a better one: A knife that breaks 20% of the time when you try to cut pork, because you didn't buy a separate pork-cutting knife. I'm not doing anything nearly as special and always have dates import as numbers for whatever reason.…

How a knife that breaks 0.00001% of the times because it’s used a billion times a day? If excel broke 20% of the time, I’d agree. But it rarely breaks. It’s just widely used. I’ve used Excel for decades. I just set the data types on my columns. The reason Excel does that is because the vast majority of people like it and rely on it. And changing it now will break millions of workflows. People assume their workflow is…

What are you talking about? You responded to the wrong person because this has nothing to do with what I said.

Re: One in five genetics papers contains errors thanks to Excel (2016)

#257

Post by thaumasiotes »

LegionMammal978 wrote: post #115
thaumasiotes wrote: post #112

Earlier quoted context omitted.

> Regardless of relative directions around circles, the sine and cosine between 0° and 90° can be described unambiguously in terms of ratios between side lengths of right triangles. That's true, if there are no angles greater than 90° or less than 0°, as is the case in a non-pathological right triangle. In this case, as ratios of nonnegative lengths, all trig functions are always nonnegative. If you want to include a…

> If you want to include angles outside those bounds, then you care about what exactly occurs where, and while you can unambiguously define angles between 0 and 90 to have all positive trig functions, you can also unambiguously define them to have negative sines and tangents. You could define trig functions differently, but then you'd need a separate pair of unnamed functions to express "the ratios of unsigned side l…

> What do you mean? "sin(0) = 0, cos(0) = 1, and for all x, sin'(x) = cos(x), cos'(x) = -sin(x)" is perfectly unambiguous.

It's pretty straightforward. "sin(0) = 0" is not a differential equation. Any phase shift applied to sine and cosine will produce exactly the same set of differential equations that apply to sine and cosine; you can rename the shifted functions "sine" and "cosine" and you'll be fine.

Re: One in five genetics papers contains errors thanks to Excel (2016)

#258

Post by kqr »

kevin_thibedeau wrote: post #245
kstrauser wrote: post #217

Earlier quoted context omitted.

I dug out my HP 50g over the weekend to play with, and conveniently… Sigh. I understand why we commonly enter math on basically a teletype-with-ASCII, and I don’t have an urge to go all APL, but for a while we were so close to a future where we could’ve had separate negation or multiplication or exponentiation symbols that might’ve removed so much room for error. I mean, that little calculator and its predecessors we…

It would be nice if OSs defaulted numeric keypad / * and - to Unicode ÷, ×, −. I never use them even when I do use the digits. That would solve the more glaring typewriter legacies. Then you'd just have the apostrophe/single-quote as the last remaining unification.

I'm not convinced. I was brought up with the middle dot for multiplication (and × reserved for cross products, I suppose?) and according to Wikipedia, the

> ISO 80000-2 standard for mathematical notation recommends only the solidus / or "fraction bar" for division, or the "colon" : for ratios; it says that the ÷ sign "should not be used" for division

I think these things are way less standardised even on paper than you believe.

Re: One in five genetics papers contains errors thanks to Excel (2016)

#259

Post by tssva »

XlA5vEKsMISoIln wrote: post #252
tssva wrote: post #250

Earlier quoted context omitted.

Excel doesn't change CSV files when it imports them. If the imported file was being changed then the user was saving back to the same file they imported from.

This information doesn't help anyone. The fact is the person was double-clicking a file in a list to view its contents and Excel was trampling it. Nobody in their right mind will waste time to open Excel first, use import feature, re-navigate to the file they were already looking at, and go through the import dialog just to see what's inside .

Trampling it to me implies that Excel was somehow modifying the contents of the file. Which it doesn't do by double clicking on the file and just viewing it. Do you mean that the data shown in Excel wasn't what was expected because of the auto data conversion?

Re: One in five genetics papers contains errors thanks to Excel (2016)

#260

Post by prepend »

the_gorilla wrote: post #256
prepend wrote: post #177

Earlier quoted context omitted.

How a knife that breaks 0.00001% of the times because it’s used a billion times a day? If excel broke 20% of the time, I’d agree. But it rarely breaks. It’s just widely used. I’ve used Excel for decades. I just set the data types on my columns. The reason Excel does that is because the vast majority of people like it and rely on it. And changing it now will break millions of workflows. People assume their workflow is…

What are you talking about? You responded to the wrong person because this has nothing to do with what I said.

> Also that's a dumb analogy. Here's a better one: A knife that breaks 20% of the time when you try to cut pork, because you didn't buy a separate pork-cutting knife.

Excel doesn’t break 20% of the time. It rarely breaks. I think you’re assuming that genetics is more worthwhile than the millions of other uses. Think about how widely it’s used and your analogy doesn’t work very well.

Post reply on HN