Hfrxdjt

I'm using Jupyter notebook working on the Kaggle Titanic dataset, trying to create a seaborn distplot to compare the 'Sex' variable (0 is male, 1 is female) with the 'Survived' variable (Survived is the project's overall dependent y-variable: 0 for did not survive, 1 for survived).

Here's my code:

fig = plt.figure(figsize=(10,10))
sns.distplot(df_bin.loc[df_bin['Survived'] == 1]['Sex'], kde_kws='label': 'Survived')
sns.distplot(df_bin.loc[df_bin['Survived'] == 0]['Sex'], kde_kws='label': 'Did not survive')

df_bin is a modified version of the training dataframe where I bin continuous variables and enumerate categorical variables. (df_bin.head() looks like this where Survived is 2nd column, Sex is 4th column):

 Survived Pclass Sex
 0 0 3 0
 1 1 1 1
 2 1 3 1
 3 1 1 1
 4 0 3 0

For 'Survived' label, the distplot looks correct, giving me a bar graph of how many survived, with a frequency line (I don't know what that line is officially called) passing through the data. But for the 'Did not survive' label, the bar chart part of the distplot looks correct, but the frequency line bounces up and down erratically--it's not smooth like the 'survived' ones.

(See picture of problem graph below for reference. I also include the correct graph from the guide I'm following below that--it uses the same code as I did so I'm confused why mine doesn't work?).

How do I get the 'Did not survive' line to stop bouncing like this and be smooth like the 'Survived' line?

Thanks,

Greg

P.S. Can someone please explain why those frequency lines are lower than the bar graph height? I would've thought they should go as high as their corresponding bar graph.