计算Python字典的百分比

Question

I want to calculate the percentage of the occurrences of the values in a dictionary. d.values/sum(d.values)*100 it gives error: unsupported operand type(s) for /: 'list' and 'int' you cannot divide the whole list with any integer. I think, and I tried with d =(Counter([Counter(i)['1'] for i in f.readlines()])) at the time of counting you can calculate the percentage but it didn't work. If anyone has an idea, please let me known.

from collections import Counter
import numpy as np
import matplotlib.pyplot as plt
import pylab as pl
with open("data_binary.txt") as f:
    d=(Counter([Counter(i)['1'] for i in  f.readlines()]))
    print d

    p = d.values()
    X = np.arange(len(d))
    pl.bar(X, p, align='center',width=0.25)
    a = np.array(d.items())
    pl.xticks(X,d.keys())
    a = a[np.argsort(a[:,0])]
    #print a
    ymax=max(d.values())+1
    pl.ylim(0, 70000)
    plt.xlabel ("Number of ones")
    plt.ylabel ("Number of Signatures")
    plt.title("Adder@1")
    pl.show()

Answer 1

You probably want to use len(d.values()) to get the number of items, unless I'm missing something. Can you please share what the dictionary (d) looks like? (ie the output of print d?)

Answer 2

you can use list comp.

In[2]: my_dict = {key: key for key in range(5)}
In[3]: values = my_dict.values()
In[4]: values
Out[4]: [0, 1, 2, 3, 4]
In[5]: total = sum(values)
In[6]: new = [value * 100. / total for value in values]
In[7]: new
Out[7]: [0.0, 10.0, 20.0, 30.0, 40.0]

or you can use np.array

In[8]: import numpy as np
In[9]: x = np.array([1, 2, 3, 4])
In[10]: x
Out[10]: array([1, 2, 3, 4])
In[11]: x/3.4
Out[11]: array([ 0.29411765,  0.58823529,  0.88235294,  1.17647059])

Answer 3

For any dictionary d, whose values are integers, the percentage of the total represented by each item can be calculated and printed as follows:

s = sum(d.values())
for k, v in d.items():
    pct = v * 100.0 / s
    print(k, pct)

numpy is overkill for this, and it's optimized for arrays rather than dictionaries.

Answer 4

Use numpy

x = np.array(d.values())
print x*100.0/sum(x)