Merging a 10x10 data table into a 1x100

[SMOK1N-ACES]

This seems like a pretty straight forward problem but the solution is escaping me... I am trying to take a data table of 10 rows and 10 columns and merge it into a 1 column with all of the 100 values... can someone post some starter code to do this? Big Thanks!

[oleksiyskononenko]

Just use dt.rbind():

from datatable import dt
DT_10x10 = dt.Frame([range(10)]*10)
DT_100x1 = dt.rbind(*DT_10x10, bynames=False)

[SMOK1N-ACES]

Thanks for the easy fix. I was struggling to implement rbind correctly.

[oleksiyskononenko]

What about simply filtering some rows out?

DT = dt.Frame(["a", "b", "c", "b"])
DT1 = DT[f.C0 != "b", :]
print(DT1)

Output

   | C0   
   | str32
-- + -----
 0 | a    
 1 | c    
[2 rows x 1 column]

[SMOK1N-ACES]

does this work for substrings tho?

I've tried to break down where I am running into road blocks but it looks like I am just digging myself a bigger hole in regards to the bigger problem I am trying to solve.... here it is in brevity...

I am trying to take a datatable and add the column names to the individual values. From there I am trying to append every column to the first column so that the resulting datatable is only 1 column. However, in the resulting single column I don't want to include any values that originally were NA or an empty string " ".

Datatable is proving to be an incredible tool for what I am doing but I am struggling a bit with implementing the correct syntax.

[oleksiyskononenko]

Yes, same approach will work for substrings. If you need to do filtering based on a particular part of the string, just use .slice(). If you need to regmatch something, take a look at .match().

[SMOK1N-ACES]

    for col in DT.names:
        DT[col] = DT[:, ifelse(col != None, col + " / " + f[col]), pass, ifelse(col != "", col + " / " + f[col], pass)]

hate to keep hammering this but the syntax seems to be escaping me... the goal here is to only create a new string if the value is NOT NA or an empty string ""....

Is there some documentation to the syntax that I should try and ingest? In my head I cant help but feel like I am tackling this from a python point of view.

[oleksiyskononenko]

Just do

for col in DT.names:
        DT[col] = DT[:, ifelse((f[col] == "") | (f[col] == None), f[col], col + " / " + f[col])]

All the documentation you need here is just ifelse.