Conversation

coding shouldnt be this hard, it's unreasonable. i asked ppl 'how many ppl have u had sex with', fill-in-the-blank and all i want to do is delete everything that isnt within a specific numerical range in the dataframe's column You'd think this would be pretty simple. you'd think.

344

Replying to

df.loc[(df["num_partners"] > x) & (df["num_partners"] < y)] This should make a new dataframe you will need to assign with only the values in the correct range.

Aella

@Aella_Girl

Replying to

@BryceDowning2

wont this not work if theres non-numerical values

10:57 PM · Oct 31, 2022

Replying to

and

df.dir = pd.to_numeric(df.dir, 'coerce') df.dir[df.dir == df.dir // 1]

Replying to

and

It's mostly about learning what to Google. Came up as first result for me for "pandas drop non integer rows"

Show replies

Replying to

Right but you can cast the values in the series to in by calling df["col_name"].to_numeric() or .astype(int)

Replying to

and

Were the answers of the form (for example): 0 1 >5 <10?

Replying to

and

Try this: df.loc[~df['num_partners'].astype(str).isnumeric(), 'num_partners'] = np.nan # Assigns NaN to non numeric values df.loc[(df['num_partners'] < 10) | (df['num_partners'] > 20), 'num_partners'] = np.nan # Assigns Nan to the out of range numeric values

Replying to

and

Sounds like you need to clean your data, your expecting code to automagically handle inconsistent information, so yeah, it's not going to be simple.

Replying to

and

Need to normalize the data before you run jobs on it. Lots of good articles on data cleaning and standardizing inputs. If you can convert all your data to date times, you can easily compare them

Show more replies