How to plot simple plot from DataFrame in Python Pandas?

user_14906835 0
Viewed 76

Hellow I have DataFrame like below:

df= pd.DataFrame({"target1" : [10, 0, 15, 10], "target2" : [50, 0, 20, 0], "ID" : ["1", "2", "3", "4"]})

And I need to create plot (bar or pie) which will show how % of ID have target1, how many target2, how many other (neither target1 nor target2).

I need results something like below:
T1 75% because 3 from 4 IDs have target1
T2 50% because 2 from 4 IDs have target2
other 25% because 1 from 4 do not have neither target1 nor target2

And I need have percentage description of columns and legend some if possible

enter image description here

pandas matplotlib plot python
2 Answers
  • set_index as 'ID' as we do not want to calculate on that column
  • Check number of items not equal to 0 (df.ne(0))
  • get mean and multiply by 100 to convert to percentage.
  • create bar plot, use rot=1, otherwise xticks would be vertical.

To annotate the bars:

  • save the plot object, loop through the patches
  • get_height (the percentage values), and format it into appropriate label, by adding '%' sign
  • get_x position and get_height, scale them by a factor marginally greater than 1, so that the labels do not intersect the bars.
>>> ax = df.set_index('ID').ne(0).mean().mul(100).plot(kind='bar', rot=1)
>>> for p in ax.patches:
        ax.annotate(str(p.get_height()) + ' %', (p.get_x() * 1.005, p.get_height() * 1.005))
>>> ax.figure

Output:

enter image description here

user_54317910

My solution

import matplotlib.pyplot as plt
import pandas as pd

df = pd.DataFrame({"target1": [10, 0, 15, 10], "target2": [50, 0, 20, 0], "ID": ["1", "2", "3", "4"]})
all = len(df["ID"])
value = [0, 0, 0]
for i in range(0, all):
    b = False
    if df["target1"][i] != 0:
        value[0] += 1
        b = True
    if df["target2"][i] != 0:
        value[1] += 1
        b = True
    if not b:
        value[2] += 1

# or you can use this
# value = [float(len([col for col in df["target1"] if col > 0])) / all * 100,
#            float(len([col for col in df["target2"] if col > 0])) / all * 100]
value = [(float(col) / all * 100) for col in value]
num_list = value
plt.bar(range(len(num_list)), num_list,
        tick_label=["target1 " + str(value[0]) + "%",
                    "target2 " + str(value[1]) + "%",
                    "default " + str(value[2]) + "%"])
plt.show()

effect

user_146462260
Related