在x轴上绘制带标签的直方图matplotlib而不是计数

Question

I'm looking to plot a histogram using value_counts() or some equivalent, in python. My data looks like:

Lannister                      8
Stark                          8
Greyjoy                        7
Baratheon                      6
Frey                           2
Bolton                         2
Bracken                        1
Brave Companions               1
Darry                          1
Brotherhood without Banners    1
Free folk                      1
Name: attacker_1, dtype: int64

You could use any reproducible code like:

pd.DataFrame({'Family':['Lannister', 'Stark'], 'Battles':[6, 8]})

I'm using

plt.hist(battles.attacker_1.value_counts())

I would like the x axis to show the family names, instead of the number of battles, and I would like the number of battles to be the histogram piece. I tried using just a series of the family names (with Lannister repeating 8 times) instead of using value_counts() and thought that might work, but I'm not sure how else to do this.

Answer 1

弄清楚了。

battles.attacker_1.value_counts().plot(kind = 'bar')

Answer 2

For a vanilla matplotlib solution, use xticklabels with xticks :

import random
import matplotlib.pyplot as plt


NUM_FAMILIES = 10

# set the random seed (for reproducibility)
random.seed(42)

# setup the plot
fig, ax = plt.subplots()

# generate some random data
x = [random.randint(0, 5) for x in range(NUM_FAMILIES)]

# create the histogram
ax.hist(x, align='left') # `align='left'` is used to center the labels

# now, define the ticks (i.e. locations where the labels will be plotted)
xticks = [i for i in range(NUM_FAMILIES)]

# also define the labels we'll use (note this MUST have the same size as `xticks`!)
xtick_labels = ['Family-%d' % (f+1) for f in range(NUM_FAMILIES)]

# add the ticks and labels to the plot
ax.set_xticks(xticks)
ax.set_xticklabels(xtick_labels)

plt.show()

Which yields:

Answer 3

You may look at pandas plot

df.set_index('Family').Battles.plot(kind='bar')