Unsorted and after Sorting Pandas with plot example¶

Based on https://stackoverflow.com/questions/44885933/how-to-sort-bars-in-a-bar-plot-in-ascending-order .

Works in notebooks launched from here.

In [1]:

#%matplotlib inline
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns

#"For examples that use the StringIO class, make sure you import it according to your Python version, i.e. from StringIO import StringIO for Python 2 and from io import StringIO for Python 3." - http://pandas.pydata.org/pandas-docs/stable/io.html
#from StringIO import StringIO # see http://pandas.pydata.org/pandas-docs/stable/io.html
from io import StringIO # see http://pandas.pydata.org/pandas-docs/stable/io.html


input ='''
    Id         Speed
    1          30
    1          35 
    1          31
    2          20
    2          25
    3          80
''' 

First the unsorted example¶

In [2]:

df = pd.read_table(StringIO(input), header=0, index_col=None,  delim_whitespace=True)
print(df.head())
result = df.groupby(["Id"])['Speed'].aggregate(np.median).reset_index()

norm = plt.Normalize(df["Speed"].values.min(), df["Speed"].values.max())
#colors = plt.cm.Reds(norm(df["Speed"])) # was causing issues when not at tmpnb.org
# and I didn't want to specify a Seaborn only palette, so just let go to default now

plt.figure(figsize=(12,8))
sns.barplot(x="Id", y="Speed", data=df) # formerly: sns.barplot(x="Id", y="Speed", data=df, palette=colors)
plt.ylabel('Speed', fontsize=12)
plt.xlabel('Id', fontsize=12)
plt.xticks(rotation='vertical')
plt.show()

   Id  Speed
0   1     30
1   1     35
2   1     31
3   2     20
4   2     25

The Result Sorted by Speed¶

In [3]:

result = df.groupby(["Id"])['Speed'].aggregate(np.median).reset_index().sort_values('Speed')
sns.barplot(x='Id', y="Speed", data=df, order=result['Id']) #formerly: sns.barplot(x='Id', y="Speed", data=df, palette=colors, order=result['Id'])
plt.show()

In [4]:

result

Out[4]:

	Id	Speed
1	2	22.5
0	1	31.0
2	3	80.0

In [ ]: