Denormalizing a column to a boolean matrix in Pandas?

Question

I'm trying to take a column of values such as:

name  tag
   a    1
   a    2
   b    2
   c    1
   b    3

and ascribe a boolean matrix with new columns, "tag_(val)", such as:

name  tag_1  tag_2  tag_3
   a    T       T      F
   b    F       T      T
   c    T       F      F

How can this be done in Pandas?

Answer 1

You could add a column full of True and then pivot:

>>> df["val"] = True
>>> piv = df.pivot("name", "tag", "val").fillna(False)
>>> piv
tag       1      2      3
name                     
a      True   True  False
b     False   True   True
c      True  False  False

Or if you're very particular:

>>> piv.columns = ["tag_{}".format(x) for x in piv.columns]
>>> piv = piv.reset_index()
>>> piv
  name  tag_1  tag_2  tag_3
0    a   True   True  False
1    b  False   True   True
2    c   True  False  False