About the Rules & Moderation category (Part 1)

Announcements Rules & Moderation
#3642

all i ask is to be able to replace the NAs in a column with 0s. but you can’t use the command that lets you replace values on a case by case basis because NA doesn’t equal NA. R should be thrown in a fire

1 Like
#3643

nyaaaaaaaaaa if N/A is NaN that is normal I think

1 Like
#3644

maybe they should consider not having this be the normal for the FUCKING DATA SCIENCE LANGUAGE WHERE YOU WILL OFTEN WANT TO REPLACE NA, ESPECIALLY EN MASSE

1 Like
#3645

that is admittedly unintuitive if you’re not aware of it though

1 Like
#3646

you’re allowed to take creative liberties with cs rules because you’re designing a new language specifically to be more useful in data science. that’s the point of designing a new language

1 Like
#3647

i’m 90% sure that the most R most people in industry actually know is the command that lets you use python

2 Likes
#3648

NYAAAAAAAAAA PYTHON BEST LANGUAGE

1 Like
#3649

(it does have the same thing with nans tho)

1 Like
#3650

note: the NA thing would not be a problem if the ifelse didn’t spit out NA upon encountering one (1) measly NA in my or statement.

1 Like
#3651

i just want to make a command that if a specific string shows up in 1 of 3 columns i get a 1 in my new column, and otherwise i get a 0. this is a simple ifelse… until you account for the fact that if that string isn’t in any of the columns and NA is it gives you NA. war.

1 Like
#3652

do you mean at least 1 of 3
or strictly 1 in 3

1 Like
#3653

something something python or also gives you the last value if all the other ones are false and the third column is mostly NA. i don’t care. it’s the data science language be useful for me

1 Like
#3654

at least. but in practice it will never show up in more than 1

1 Like
#3655

does N/A == N/A return N/A?? instead of False???

1 Like
#3656

you don’t have to do this work btw. since i already did it and just found the scuffed command to after the fact change the NAs to 0’s. which is very dumb

1 Like
#3657

R

1 Like
#3658

nyaaaaaaa I realize
but I was wondering if there was a way to make it clean with shortcircuiting

1 Like
#3659

this is when an AND or OR statement already knows the result (because the first condition was FALSE for AND or TRUE for OR), so it doesn’t evaluate the second condition at all

1 Like
#3660

besides speeding up the code, it’s useful for clean null-checks

ex:
if (obj is not None and obj.attribute_name > 1) isn’t going to give you an error if obj is None, because if obj is None the second part is never evaluated (putting the conditions the other way around would give the error)

1 Like
#3661

fun fact: the command that lets you change column values en masse is also dumb! because it also forces you to include even values you don’t plan on changing, just to assert that they are in fact the same. and anything you miss becomes NA (which, of course, you can’t change with that command because NA != NA).

1 Like