Show accuracy

[ntdgo]

Hi, I'm used KNN for my research but I don't know how to display accuracy of a result
This is my code:

# In[38]:
from sklearn import neighbors, metrics
from sklearn.neighbors import NearestNeighbors
import numpy as np
import pandas as pd
import sys
import json
import math

data    = pd.read_excel('dataset.xlsx')
data    = np.array(data.as_matrix())
# In[40]:
knn=neighbors.KNeighborsClassifier(n_neighbors=5)
# In[41]:
X = data[:,:-3]
Y = data[:,-1:]
Y = np.zeros(len(Y))
for i in range(0,len(Y)):
    if data[i,5] == 0:
        Y[i] = 0
    elif data[i,5] == 1:
        Y[i] = 1
    elif data[i,5] == 2:
        Y[i] = 2
    elif data[i,5] == 3:
        Y[i] = 3
    elif data[i,5] == 4:
        Y[i] = 4
# In[42]:
knn.fit(X, Y)
# In[49]:
result = knn.predict([[220.4, 6.39,1855]])
print(result)
# result = knn.predict(X)
# print(metrics.accuracy_score(Y[2000], result))

This is my dataset:

Thank you!

[abhaymise]

Can you elaborate your problem a little.
As per your code you are trying to get prediction for a sample set.
You can get proability score of the sample by calling predict_proba(X).

To get mean accuracy for a heldout data you can call score(X, y,)

[ntdgo]

Something wrong here, I get predict_proba from "sklearn.svm.libsvm." but it has to at least 6 arguments, could you help me!

[justmarkham]

@CrackerNo1 It seems like you could benefit from watching my video series:
https://www.youtube.com/playlist?list=PL5-da3qGB5ICeMbQuqbbCOQWcS6OYBr5A