Unable to use seaborn.countplot - pandas

Question

I'm trying to plot some graphs using the latest version of Pycharm as a Python IDE.
As an interpreter, I'm using Anaconda with Python 3.4.3-0.
I have installed using conda install the news version of pandas (0.17.0), seaborn (0.6.0), numpy (1.10.1), matplotlib (1.4.3), ipython (4.0.1)
Inside the nesarc_pds.csv I have this:
IDNUM,S1Q2I
39191,1
39787,1
40082,1
40189,1
40226,1
40637,1
41306,1
41627,1
41710,1
42113,1
42120,1
42720,1
42909,1
43092,1
7,2
15,2
25,2
40,2
46,2
49,2
57,2
63,2
68,2
100,2
104,2
116,2
125,2
136,2
137,2
145,2
168,2
3787,9
6554,9
7616,9
11686,9
12431,9
14889,9
17694,9
19440,9
20141,9
21540,9
22476,9
24207,9
25762,9
29045,9
29731,9
So, that being said, this is my code:
import pandas as pd
import numpy
import seaborn as snb
import matplotlib.pyplot as plt
data = pd.read_csv("nesarc_pds.csv", low_memory=False)
#converting variable to numeric
pd.to_numeric(data["S1Q2I"], errors='coerce')
#setting a new dataset...
sub1=data[(data["S1Q2I"]==1) & (data["S3BQ1A5"]==1)]
sub2 = sub1.copy()
#setting the missing data 9 = unknown into NaN
sub2["S1Q2I"] = sub2["S1Q2I"].replace(9, numpy.nan)
#setting date to categorical type
sub2["S1Q2I"] = sub2["S1Q2I"].astype('category')
#plotting
snb.countplot(x="S1Q2I", data=sub2)
plt.xlabel("blablabla")
plt.title("lalala")
And then.....this is the error:
Traceback (most recent call last):
File "C:/Users/LPForGE_1/PycharmProjects/guido/haha.py", line 49, in <module>
snb.countplot(x="S1Q2I", data=sub2)
File "C:\Anaconda3\lib\site-packages\seaborn\categorical.py", line 2544, in countplot
errcolor)
File "C:\Anaconda3\lib\site-packages\seaborn\categorical.py", line 1263, in __init__
self.establish_colors(color, palette, saturation)
File "C:\Anaconda3\lib\site-packages\seaborn\categorical.py", line 300, in establish_colors
l = min(light_vals) * .6
ValueError: min() arg is an empty sequence
Any help would be really nice. I pretty much exhausted my intelligence trying to understand how to solve this.