>rounding errors

>>8381294
Use double instead of float. Or throw your calculator into the trash.

>>838129
Or just round and deal with it sissy

are the reason many man-made systems are collapsing

>>8381294
>double """precision""" IEEE754

>>8381294
>he doesn't precondition

File: 1430800891931.jpg (10KB, 267x200px) Image search: [Google]

10KB, 267x200px

>go to university from a community college
>Professors are more dickish than usual

>>8381294
This is why the sciences are fucking trash.

>Don't forget to keep track of your sig figs :^)
Fuck off. Thanks god you don't have to deal with this gay shit in pure mathematics.

>>8381294
>he doesn't do all his calculations in a way to ensure arbitrary precision

>>8383341
>i hate reality
t. schizo

>>8383341

On the one hand, as a math grad, I appreciate your sentiment.

On the other, mathematics is a science, or if you're going to get shirty about that, then mathematics is certainly in "the sciences". And if you would object that pure math entails the pleasure of purely autistic exercises for their own sake (which have little to no hope of application), as opposed to "advancing the human condition" or whatever practical purpose "the sciences" are supposed to have, then it is worth pointing out that the rest of the sciences have their own autistic exercises (large hard-on colliders) which do not yet (but may someday) have some happy application. The point being that math is like the rest of the sciences in that the people who do them like to do it for its own sake to some degree, and intelligent society seems to understand that there is value in spending money on pure R&D, even if a lot of it was initially motivated by military application, or if scads of the stuff is only of interest to the scientists themselves for the space of a given century.

>>8383352

I'm not that anon, but your implication that reality is not contemptible, ought to be "accepted" on some level for its own sake, and if it isn't then the person who has a problem with reality is somehow damaged, is itself both depressing and laughable.

>>8383355
>I'm not that anon!! *spouts autistic crap*
you are that anon

>>8383491
But he's not... I am.

>>8381294
Ever use Microsoft Excel for science shit?

It automatically changes some shorthand notations into dates and long numbers into scientific notations. You can't turn the feature off. Microsoft has known about it for like 15 years and refuses to allow users to turn it off. Data for tons of scientific experiments are ruined because of this.

https://www.washingtonpost.com/news/wonk/wp/2016/08/26/an-alarming-number-of-scientific-papers-contain-excel-errors/

File: 1475019775241.png (123KB, 504x416px) Image search: [Google]

123KB, 504x416px

>>8383928
and every single scientist that fucks up their data because of that deserves it. I am not defending microsoft, but from personal experience I have lost all sympathy for a "scientist" that thinks it is below them to learn proper tools for their craft.

File: IMG_2111.png (102KB, 258x245px) Image search: [Google]

102KB, 258x245px

>>8383928
>Excel
>science

ok

File: 1471196325250.gif (2MB, 209x229px) Image search: [Google]

2MB, 209x229px

>>8384032
no, people actually do that. I've seen chemists and biologists do it alike.

>>8383888
Stop posing as me faggot.

>>8384029
>>8384032
>>8384037

What do scientists use instead of Excel...?

please don't bully me i'm a mathfag

>>8384072
R

>>8384072
>mathfag
>doesn't use Matlab
>pls

>>8383341
Sig figs are important though, rounding too soon can give you an answer as much as 10% off

>>8384130
or as much as 100%

>>8384130
I remember having to take basic chem in uni just as a requirement of the engineering department (EECS @ Berkeley), holy fuck my professor was weird about sig figs. I was used to only rounding at the end of a problem from physics and practically every other math/science course, but this dude wanted rounding at each step + using the rounded answer from previous parts in multipart problems. Made checking answers right after an exam really weird, even basic empirical/molecular formulas would never match up.

>>8384032
>>8384029
>>8384037
Some are not allowed to use anything else. Their employers dictate what tools they use in those cases. Which is fucking absurd obviously. A lot of government research is like that. You absolutely can not use anything not fully approved and getting shit approved means being neck deep in paper work and kissing assholes.

File: 1473648988582.jpg (178KB, 743x960px) Image search: [Google]

178KB, 743x960px

>>8384143
>rounding at each step

What a colossal fucking tard.

>>8384032
You know whats the fucking worst thing about excel.

It automatically converts number to numbers.

We had to use it at work to edit some csv files that contains long number strings like '46874564354867654654872103218'. Just opening it in excell and saving the file again saves them all as 4.68x10^20. So getting the backup and trying again, we open, edit, then before saving set the columns to text, save, now its 46874564354867650000000000000. It fucking drives me nuts. In the end I did it with a python script at home (because we cant use anything else on work computers).

I cant balieve they keep using it.

>>8384151
you're right some don't get to decide, I keep forgetting that. I'm sorry for those people, such systems breed bad science in general.

>>8384155
this, only round at the end

>>8384155
>>8384143
I had a professor once that did something similar, I think he just liked screwing with people because we had to round after each 'answer', but not in the middle of a calculation. even if the 'answer' was in the middle of a question. Everyone lost about 20% each test because they rounded in the wrong places.

Example: Lest say you have to calculate The flow rate of substance A, and to get it you first need to get the flow rate of substance B and the temperature.

>Calculate flow rate of substance B <<Dont round in the process of doing this
>Round the flow rate of substance B to 2 decimals
>Calculate temperature from flow rate B <<Dont round in the process of doing this BUT use the rounded flow rate of B
>Round the final temperature to 2 decimals
>Calculate flow rate A <<Use rounded temperature and flow rate B

And added to this, he didnt care about significant figures, he only cared about the decimals, so if some answer was 4516.846mm and you have to round it to 2 decimal places you should should OBVIOUSLY round it to 4510mm because the professors answer was in meter so its 4.5168m ~ 4.51m = 4510mm.