python - SettingWithCopyWarning while using .loc -

problem simplified:

i need extract , modify particular rows of dataframe based on whether or not text within column has '-' character. dash , beyond needs removed , remaining text needs whatever preceding '-'.

have:      textcol 0    no dash here 1    1 - here  want:      textcol 0    1 

here code used recreate scenario.

df = pd.dataframe(data=['no dash here', 'one - here'], index=[0, 1], columns=['textcol']) df2 = df[df['textcol'].str.contains('-') == true] df2.loc[:, ['textcol']] = df2['textcol'].str.split('-').str[0] 

the resulting dataframe df2 yields result desire, 1 exception. every time call df2 (or derivative thereafter) receive following settingwithcopywarning:

a value trying set on copy of slice dataframe see caveats in documentation: http://pandas.pydata.org/pandas-docs/stable/indexing.html#indexing-view-versus-copy 

i tried accomplish wanted different way, , given similar error instructed me try , use .loc() functionality instead, i'm still receiving similar error.

is there better, non-error threatening way me accomplish result? i'm afraid occurring here don't understand , df2 not result in want. wondering if .query() work.

as mentioned @edchum, df2 view on df opposed copy. if want copy, can use .copy() (see docs) , settingwithcopywarning disappears:

df2 = df[df['textcol'].str.contains('-') == true].copy() df2.loc[:, ['textcol']] = df2['textcol'].str.split('-').str[0] 

see returning view vs copy in pandas docs.