Can we sort the x axis in a Python plot after plotting it?

Question

I am trying to plot two columns from two different dataframes in one graph. I have figured out the code to plot it in one but the x axis needs to be in ascending order. Can we do that in this case?

df1=pd.DataFrame({ 'ID': ['A1','A2','A6','A7','A9'], 'F1_S': [23,75,42,77,54] },
                 columns=['ID', 'F1_S'])
df2=pd.DataFrame({ 'ID': ['A3','A4','A5','A8'], 'F1_S': [66,43,56,86] },
                 columns=['ID', 'F1_S'])
x1 = list(df1['ID'])
y1= list(df1['F1_S'])
lists1 = sorted(zip(*[x, y1]))
x1, y1 = zip(*sorted(zip(x1, y1)))

x2 = list(df2['ID'])
y2= list(df2['F1_S'])
lists2 = sorted(zip(*[x2, y2]))
x2, y2 = zip(*sorted(zip(x2, y2)))

plt.plot (x1,y1)
plt.plot(x2,y2)

@sobek but the axis labels aren't in order, look more closely. also, there's an A9 that should come after A8 — wpercy
– wpercy, Commented Sep 6, 2018 at 19:58

ImportanceOfBeingErnest · Accepted Answer · 2018-09-06 20:42:25Z

One possible solution is to plot the concatenated dataframe against a numeric index and set the ticks to categories.

import matplotlib.pyplot as plt
import pandas as pd

df1=pd.DataFrame({ 'ID': ['A1','A2','A6','A7','A9'], 'F1_S': [23,75,42,77,54] },
                 columns=['ID', 'F1_S'])
df2=pd.DataFrame({ 'ID': ['A3','A4','A5','A8'], 'F1_S': [66,43,56,86] },
                 columns=['ID', 'F1_S'])
# Rename the column of one of the dataframe
df2 = df2.rename(columns={'F1_S' : 'F2_S'})
# Concatenate dataframes, reset the index (such that it will be numeric)
df3 = pd.concat([df1.set_index("ID"), df2.set_index("ID")], sort=True).sort_index().reset_index()

#Plot each column without nan values, against the numeric index
plt.plot(df3["F1_S"].dropna())
plt.plot(df3["F2_S"].dropna())
# Set labels to category names
plt.xticks(range(len(df3)), df3["ID"])
plt.show()

sobek · Accepted Answer · 2018-09-06 21:20:23Z

Not sure if this is entirely what you want but here goes:

import pandas as pd
import matplotlib.pyplot as plt


df1 = pd.DataFrame({'ID': ['A1', 'A2', 'A6', 'A7', 'A9'],
                    'F1_S': [23, 75, 42, 77, 54]},
                   columns=['ID', 'F1_S'])
df2 = pd.DataFrame({'ID': ['A3', 'A4', 'A5', 'A8'],
                    'F1_S': [66, 43, 56, 86]},
                   columns=['ID', 'F1_S'])

df1.set_index('ID', inplace=True)
df1.index.name = None
df1.columns = ['F1_S_1']
df2.set_index('ID', inplace=True)
df2.index.name = None
df2.columns = ['F1_S_2']
df = pd.concat([df1, df2], sort=True).sort_index().interpolate(limit_area='inside')
ax = df.plot()
ax.set_xticklabels(['A0'] + df.index.tolist())
plt.show()

The result is this:

df.plot() seems to have an off-by-one error with the axis labeling, i had to hack a fake element into the xlabel list to get it to display correctly.

PMende · Accepted Answer · 2018-09-06 20:05:25Z

0

Why not just concatenate the two?

pd.concat([df1[['ID', 'F1_S']], df2[['ID', 'F1_S']])\
  .sort_values(by='ID')\
  .plot(x='ID', y='F1_S')

Edit: to make them separate lines plotted in the same axis:

fig, ax = plt.subplots()
df1.sort_values(by='ID')\
   .plot(x='ID', y='F1_S', ax=ax)
df2.sort_values(by='ID')\
   .plot(x='ID', y='F1_S', ax=ax)

The call to df.plot accepts the same kwargs asplt.plot, so you can specify different colors for each.

edited Sep 6, 2018 at 20:05

answered Sep 6, 2018 at 19:55

PMende

5,4903 gold badges21 silver badges26 bronze badges

3 Comments

ImportanceOfBeingErnest Over a year ago

This way you would get one line instead of two.

MasterShifu Over a year ago

I wanted the graph in two different colors to be able to differentiate which points belong to which dataframe

MasterShifu Over a year ago

The updated code is plotting point with x = df1['ID'], y = df2['F1_S'] and not x = df2['ID'], y = df2['F1_S']

Collectives™ on Stack Overflow

Can we sort the x axis in a Python plot after plotting it?

3 Answers 3

Comments

Comments

3 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

3 Answers 3

Comments

Comments

3 Comments

Your Answer

Sign up or log in

Post as a guest

Related