What I'd like to do is create a matlab scatter plot as below, but color the dots according to the the string in column C, for example, yes = blue and no = red.
import matplotlib.pyplot as plt
A = pd.DataFrame.from_items([('A', [1, 2, 3]), ('B', [4, 5, 6]), ('C', ['yes', 'yes', 'no'])])
plt.scatter(A['A'], A['B'])
You can create a dictionary of (value: color) pairs and use those values as color argument on plt.scatter using a list comprehension.
colors = {'yes': 'b', 'no': 'r'}
plt.scatter(A['A'], A['B'], color=[colors[r] for r in A['C']])
EDIT: Preferred way to add legend is to iterate over unique classes in your dataset and use legend argument. So we can iterate over the dataframe grouped by C and plot them individually with label arguments. And finally call plt.legend() to show it.
colors = {'yes': 'b', 'no': 'r'}
for idx, row in A.groupby('C'):
plt.scatter(row['A'], row['B'],
color=[colors[r] for r in row['C']],
label=idx)
plt.legend()
Funny one-liner:
plt.scatter(x="A",y="B",c="C",
data=df.apply(lambda x: [x.A,x.B,chr(int(1.6*ord(x.C[1])-63.6))], axis=1))
plt.scatter(A['A'], A['B'], color=['b' if r=='yes' else 'r' for r in A['C']]) for a one liner (and be boring) but that is pretty cool.