removing missing values from coloumns

[Hasham04]

col1 col2
value1 ''
'' value1
value2 ''
'' value2

any suggestions on how i can i convert this data to

col1 col2
value1 value1
value2 value2

I then want to be able to export it to a CSV.

thankyou

[Hasham04]

One approach that I tried was to filter each column so

df_col1_filtered=df_unfiltered[df_unfiltered.col1!='']
df_col2_filtered=df_unfiltered[df_unfiltered.col2!='']

when I describe this data I get an equal number of counts of rows. but when I try to join them like this.

df_col1_filtered.drop(col2,inplace=True)
df_col2_filtered.drop(col1,inplace=True)

joined=df_col1_filtered.join(df_col2_filtered)

I get an unequal number of rows being merged error.

I would appreciate any suggestions to solve the problem I mentioned in question or any insight on what I am doing wrong with my approach. Thank you for your help.

[Hasham04]

im using df.func.where(condition,val1,val2) to make the columns i mentioned above.

how do I pass na to val2 so that i can use dropna.

[maartenbreddels]

you can pass None

[Hasham04]

i have already tried that but i get

None is not a an Arrow or NumPy array

[JovanVeljanoski]

This is answered somewhere on this issue board. But why do you wanna pass NA to the output of a where statement?

[Hasham04]

i have data that looks like this

col1 col2 col3 col4
val1 val2 val3 val4

making this data look like this

with the where statement
df[newcol1]=df.func.where(col1=='something',col2,'')
df[newcol2]=df.func.where(col3=='something',col4,'')

col1 col2 col3 col4 newcol1 newcol2

I then drop col1 col2 col3 col4.

which makes the data look like this.
newcol1 newcol2
value1 ''
'' value1
value2 ''
'' value2.

i am then filtering each column right now like this

df.newcol2=df_unfiltered[df.newcol2!='']

Ii was wondering if could do

newcol1=newcol1.dropna()
so I was wondering instead of filtering out '' if I could just use dropna .