In Sci-kit learn have created a few models with train and test data.
The models work fine, but when I try to compute any accuracy metrics, it fails. I assume something is wrong with either my prediction object (pred y) or expected object (true y).
For this test, I have looked at the pred y. It is an object and have 119 0/1 values.
The true y is also an object and has 119 0/1 values.
My code and the error is below, as well as an object comparison. It is the error I do not understand.
"expected" is my true y and "target_predicted" is the predicted y.
I have tried other metrics and other models- it always fails when I am at this stage.
Any assistance?
#Basic Decsion Tree
clf = tree.DecisionTreeClassifier()
clf = clf.fit(bank_train, bank_train_target)
print clf
DecisionTreeClassifier(compute_importances=None, criterion='gini',
max_depth=None, max_features=None, max_leaf_nodes=None,
min_density=None, min_samples_leaf=1, min_samples_split=2,
random_state=None, splitter='best')
#test model using test data
target_predicted = clf.predict(bank_test)
accuracy_score(expected,target_predicted)
#error
---------------------------------------------------------------------------
ValueError Traceback (most recent call last)
<ipython-input-31-23d1a990a192> in <module>()
1 #test model using test data
2 target_predicted = clf.predict(bank_test)
----> 3 accuracy_score(expected,target_predicted)
/Users/mpgartland1/anaconda/lib/python2.7/site-packages/sklearn/metrics/metrics.pyc in accuracy_score(y_true, y_pred, normalize, sample_weight)
1295
1296 # Compute accuracy for each possible representation
-> 1297 y_type, y_true, y_pred = _check_clf_targets(y_true, y_pred)
1298 if y_type == 'multilabel-indicator':
1299 score = (y_pred != y_true).sum(axis=1) == 0
/Users/mpgartland1/anaconda/lib/python2.7/site-packages/sklearn/metrics/metrics.pyc in _check_clf_targets(y_true, y_pred)
125 if (y_type not in ["binary", "multiclass", "multilabel-indicator",
126 "multilabel-sequences"]):
--> 127 raise ValueError("{0} is not supported".format(y_type))
128
129 if y_type in ["binary", "multiclass"]:
ValueError: unknown is not supported
Here is a comparison of the two objects.
print target_predicted.size
print expected.size
print target_predicted.dtype
print expected.dtype
print target_predicted
print expected
119
119
object
object
[1 0 0 1 0 0 1 0 1 1 1 0 1 1 0 1 1 1 0 0 1 0 0 1 0 1 0 1 0 0 0 0 0 1 1 0 1
0 0 0 0 0 0 0 1 0 0 0 1 1 1 1 1 0 0 1 1 0 0 1 1 0 1 1 1 1 1 1 1 0 1 0 0 0
0 1 0 0 1 1 1 0 1 1 0 0 0 0 1 0 0 0 0 0 0 1 0 1 0 1 0 0 0 1 0 1 1 0 0 0 1
0 1 0 1 0 0 0 1]
[1 0 0 1 0 0 1 0 1 1 1 1 1 0 1 1 1 0 1 0 1 0 0 1 0 1 1 1 1 0 0 0 0 1 1 1 1
0 0 1 0 0 0 1 1 0 0 0 0 1 1 1 1 0 0 1 1 0 1 1 1 0 1 1 1 1 0 1 1 0 0 0 0 0
0 1 0 0 1 1 1 0 1 1 0 0 0 0 1 0 0 1 0 0 0 1 1 0 0 1 0 0 0 1 0 1 1 0 0 0 1
0 1 0 0 0 1 0 1]
If also fails when I try a confusion matrix or other metric- using very cookie cutter code. So, my guess is in the object(s).
Thanks
Related
I'm trying to find the confusion matrix of a multiclass classification problem. I'm using tf.math.confusion_matrix() to do that. The code snippet is as follows,
y_pred = model.predict(x_test)
y_pred = tf.argmax(y_pred, axis=1)
Y_test = tf.argmax(y_test, axis=1)
matrix = tf.math.confusion_matrix(Y_test, y_pred)
The output of Y_test is,
tf.Tensor(
[[0 2 0 ... 0 0 0]
[0 2 0 ... 0 0 0]
[0 2 0 ... 0 0 0]
...
[0 0 0 ... 0 0 0]
[0 0 3 ... 0 0 0]
[0 0 2 ... 0 0 0]], shape=(2124, 279), dtype=int64)
The output of y_pred is,
tf.Tensor(
[[1 2 2 ... 0 0 0]
[0 2 3 ... 0 0 0]
[3 2 0 ... 3 1 3]
...
[3 1 0 ... 2 3 2]
[1 0 3 ... 1 1 2]
[1 0 2 ... 1 1 2]], shape=(2124, 279), dtype=int64)
Y_test[1] looks like the following,
tf.Tensor(
[0 2 0 1 0 3 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0], shape=(279,), dtype=int64)
y_pred[1] looks like the following,
tf.Tensor(
[0 2 3 1 3 3 2 2 2 3 2 3 2 3 3 2 1 0 0 0 0 3 1 0 2 3 1 2 0 1 0 0 1 0 0 0 0
2 0 2 1 0 0 0 0 1 0 0 0 3 2 0 0 3 2 0 0 3 3 0 3 0 0 0 0 1 0 2 1 0 2 3 0 3
3 0 2 3 1 3 2 0 3 0 0 0 0 0 0 0 0 0 0 0 0 1 3 0 0 0 3 3 0 0 0 0 0 3 0 0 1
0 3 0 3 3 0 1 0 3 0 0 0 0 0 0 3 0 1 0 0 0 0 0 0 0 0 0 3 0 0 3 0 0 0 0 0 0
3 0 3 3 0 0 0 3 0 0 0 0 0 0 0 0 0 0 3 0 3 0 0 0 0 3 0 3 0 0 0 0 0 0 0 2 0
0 1 0 0 0 0 2 0 3 3 0 0 0 0 0 0 0 0 0 0 0 0 0 3 0 0 0 3 0 3 3 0 2 3 0 3 3
3 3 3 0 3 0 3 0 0 3 0 0 0 0 3 3 3 2 0 0 0 0 0 0 0 0 2 3 0 0 3 0 0 0 3 0 2
0 0 3 0 0 0 0 0 3 1 2 0 3 2 3 0 3 0 0 0], shape=(279,), dtype=int64)
And the error I'm getting is,
InvalidArgumentError: Dimensions [0,2) of indices[shape=[2124,2,279]] must match dimensions [0,2) of updates[shape=[2124,279]] [Op:ScatterNd]
How this can be solved?
I load data from a csv file. When I try to save the labels column into a variable I am getting the error saying label, and the dataset was taken from https://www.kaggle.com/c/digit-recognizer/data
d0 = pd.read_csv('./test.csv')
print(d0.head(5)) # print first five rows of d0.
# to save the labels into a variable l.
l = (d0['label'])
got output:
pixel0 pixel1 pixel2 pixel3 pixel4 pixel5 pixel6 pixel7 pixel8 \
0 0 0 0 0 0 0 0 0 0
1 0 0 0 0 0 0 0 0 0
2 0 0 0 0 0 0 0 0 0
3 0 0 0 0 0 0 0 0 0
4 0 0 0 0 0 0 0 0 0
pixel9 ... pixel774 pixel775 pixel776 pixel777 pixel778 \
0 0 ... 0 0 0 0 0
1 0 ... 0 0 0 0 0
2 0 ... 0 0 0 0 0
3 0 ... 0 0 0 0 0
4 0 ... 0 0 0 0 0
pixel779 pixel780 pixel781 pixel782 pixel783
0 0 0 0 0 0
1 0 0 0 0 0
2 0 0 0 0 0
3 0 0 0 0 0
4 0 0 0 0 0
[5 rows x 784 columns]
I'm getting error
KeyError Traceback (most recent call last)
~\Anaconda3\lib\site-packages\pandas\core\indexes\base.py in get_loc(self, key, method, tolerance)
3077 try:
-> 3078 return self._engine.get_loc(key)
3079 except KeyError:
KeyError: 'label'
During handling of the above exception, another exception occurred:
KeyError Traceback (most recent call last)
in
15
16 # save the labels into a variable l.
---> 17 l = (d0['label'])
18
19 # Drop the label feature and store the pixel data in d.
KeyError: 'label'
I cant understand the error described in anaconda
Dropped some error code as stack is showing add more detail error
Please help me on solving this problem
I am doing a News recommendation system and I need to build a table for users and news they read. my raw data just like this :
001436800277225 [12,456,157]
009092130698762 [248]
010003000431538 [361,521,83]
010156461231357 [173,67,244]
010216216021063 [203,97]
010720006581483 [86]
011199797794333 [142,12,86,411,201]
011337201765123 [123,41]
011414545455156 [62,45,621,435]
011425002581540 [341,214,286]
the first column is userID, the second column is the newsID.newsID is a index column, for example, after transformation, [12,456,157] in the first row means that this user has read the 12th, 456th and 157th news (in sparse vector, the 12th column, 456th column and 157th column are 1, while other columns have value 0). And I want to change these data into a sparse vector format that can be used as input vector in Kmeans or DBscan algorithm of sklearn.
How can I do that?
One option is to construct the sparse matrix explicitly. I often find it easier to build the matrix in COO matrix format and then cast to CSR format.
from scipy.sparse import coo_matrix
input_data = [
("001436800277225", [12,456,157]),
("009092130698762", [248]),
("010003000431538", [361,521,83]),
("010156461231357", [173,67,244])
]
NUMBER_MOVIES = 1000 # maximum index of the movies in the data
NUMBER_USERS = len(input_data) # number of users in the model
# you'll probably want to have a way to lookup the index for a given user id.
user_row_map = {}
user_row_index = 0
# structures for coo format
I,J,data = [],[],[]
for user, movies in input_data:
if user not in user_row_map:
user_row_map[user] = user_row_index
user_row_index+=1
for movie in movies:
I.append(user_row_map[user])
J.append(movie)
data.append(1) # number of times users watched the movie
# create the matrix in COO format; then cast it to CSR which is much easier to use
feature_matrix = coo_matrix((data, (I,J)), shape=(NUMBER_USERS, NUMBER_MOVIES)).tocsr()
Use MultiLabelBinarizer from sklearn.preprocessing
from sklearn.preprocessing import MultiLabelBinarizer
mlb = MultiLabelBinarizer()
pd.DataFrame(mlb.fit_transform(df.newsID), columns=mlb.classes_)
12 41 45 62 67 83 86 97 123 142 ... 244 248 286 341 361 411 435 456 521 621
0 1 0 0 0 0 0 0 0 0 0 ... 0 0 0 0 0 0 0 1 0 0
1 0 0 0 0 0 0 0 0 0 0 ... 0 1 0 0 0 0 0 0 0 0
2 0 0 0 0 0 1 0 0 0 0 ... 0 0 0 0 1 0 0 0 1 0
3 0 0 0 0 1 0 0 0 0 0 ... 1 0 0 0 0 0 0 0 0 0
4 0 0 0 0 0 0 0 1 0 0 ... 0 0 0 0 0 0 0 0 0 0
5 0 0 0 0 0 0 1 0 0 0 ... 0 0 0 0 0 0 0 0 0 0
6 1 0 0 0 0 0 1 0 0 1 ... 0 0 0 0 0 1 0 0 0 0
7 0 1 0 0 0 0 0 0 1 0 ... 0 0 0 0 0 0 0 0 0 0
8 0 0 1 1 0 0 0 0 0 0 ... 0 0 0 0 0 0 1 0 0 1
9 0 0 0 0 0 0 0 0 0 0 ... 0 0 1 1 0 0 0 0 0 0
I have pandas data frame like below.
df
Out[50]:
0 1 2 3 4 5 6 7 8 9 ... 90 91 92 93 94 95 96 97 \
0 0 0 0 0 0 0 0 0 0 0 ... 1 1 1 1 1 1 1 1
1 0 1 1 1 0 0 1 1 1 1 ... 0 0 0 0 0 0 0 0
2 1 1 1 1 1 1 1 1 1 1 ... 0 0 0 0 0 0 0 0
3 0 0 0 0 0 0 0 0 0 0 ... 1 1 1 1 1 1 1 1
4 0 0 0 0 0 0 0 0 0 0 ... 1 1 1 1 1 1 1 1
5 1 0 0 1 1 1 1 0 0 0 ... 0 0 0 0 0 0 0 0
6 0 0 0 0 0 0 0 0 0 0 ... 1 1 1 1 1 1 1 1
7 0 0 0 0 0 0 0 0 0 0 ... 1 1 1 1 1 1 1 1
[8 rows x 100 columns]
I have target variable as an array as below.
[1, -1, -1, 1, 1, -1, 1, 1]
How can I map this target variable to a data frame and convert it into lib SVM format?.
equi = {0:1, 1:-1, 2:-1,3:1,4:1,5:-1,6:1,7:1}
df["labels"] = df.index.map[(equi)]
d = df[np.setdiff1d(df.columns,['indx','labels'])]
e = df.label
dump_svmlight_file(d,e,'D:/result/smvlight2.dat')er code here
ERROR:
File "D:/spyder/april.py", line 54, in <module>
df["labels"] = df.index.map[(equi)]
TypeError: 'method' object is not subscriptable
When I use
df["labels"] = df.index.list(map[(equi)])
ERROR:
AttributeError: 'RangeIndex' object has no attribute 'list'
Please help me to solve those errors.
I think you need convert index to_series and then call map:
df["labels"] = df.index.to_series().map(equi)
Or use rename of index:
df["labels"] = df.rename(index=equi).index
All together:
For difference of columns pandas has difference:
from sklearn.datasets import dump_svmlight_file
equi = {0:1, 1:-1, 2:-1,3:1,4:1,5:-1,6:1,7:1}
df["labels"] = df.rename(index=equi).index
e = df["labels"]
d = df[df.columns.difference(['indx','labels'])]
dump_svmlight_file(d,e,'C:/result/smvlight2.dat')
Also it seems label column is not necessary:
from sklearn.datasets import dump_svmlight_file
equi = {0:1, 1:-1, 2:-1,3:1,4:1,5:-1,6:1,7:1}
e = df.rename(index=equi).index
d = df[df.columns.difference(['indx'])]
dump_svmlight_file(d,e,'C:/result/smvlight2.dat')
Excel Formulas I am trying to replicate in pandas:
Click here to download workbook
* Look at columns D, E and F
entsig and exsig are manual and can be changed. In real life they would be derived from the value of another column or a comparison of two other columns
ent = 1 if entsig previous = 1 and in = 0
in = 1 if ent previous = 1 or (in previous = 1 and ex = 0)
ex = 1 if exsig previous = 1 and in previous = 1
so either ent, in, or ex will always be = 1 but never more than one of them
import pandas as pd
df = pd.DataFrame(
[[0,0,0,0,0], [1,0,0,0,0], [1,0,0,0,0], [1,0,0,0,0], [0,0,0,0,0],
[0,1,0,0,0], [0,1,0,0,0], [1,0,0,0,0], [1,0,0,0,0], [0,0,0,0,0],
[0,0,0,0,0], [0,0,0,0,0], [0,1,0,0,0], [0,1,0,0,0], [0,1,0,0,0],
[0,0,0,0,0], [0,0,0,0,0], [1,0,0,0,0], [1,0,0,0,0], [1,0,0,0,0],
[1,1,0,0,0], [0,1,0,0,0], [0,1,0,0,0], [0,1,0,0,0]],
columns=['entsig', 'exsig','ent', 'in', 'ex'])
for i in df.index:
df['ent'][(df.entsig.shift(1)==1) & (df['ent'].shift(1) == 0) & (df['in'].shift(1) == 0)]=1
df['ex'][(df.exsig.shift(1)==1) & (df['in'].shift(1)==1)]=1
df['in'][(df.ent.shift(1)==1) | ((df['in'].shift(1)==1) & (df['ex']==0))]=1
for j in df.index:
df['ent'][df['in'] == 1]=0
df['in'][df['ex']==1]=0
df['ex'][df['ex'].shift(1)==1]=0
df
results in
entsig exsig ent in ex
0 0 0 0 0 0
1 1 0 0 0 0
2 1 0 1 0 0
3 1 0 0 1 0
4 0 0 0 1 0
5 0 1 0 1 0
6 0 1 0 0 1
7 1 0 0 0 0
8 1 0 1 0 0
9 0 0 0 1 0
10 0 0 0 1 0
11 0 0 0 1 0
12 0 1 0 1 0
13 0 1 0 0 1
14 0 1 0 0 0
15 0 0 0 0 0
16 0 0 0 0 0
17 1 0 0 0 0
18 1 0 1 0 0
19 1 0 0 1 0
20 1 1 0 1 0
21 0 1 0 0 1
22 0 1 0 0 0
23 0 1 0 0 0
Question
How can I make this code faster? It runs slow because it's a loop but I have not been able to come up with a solution that does not use loops. Any ideas or comments are appreciated.
If we can assume every group of 1's in entsig is followed by at least one 1 in
exsig, then you could compute ent, ex and in like this:
def ent_in_ex(df):
entsig_mask = (df['entsig'].diff().shift(1) == 1)
exsig_mask = (df['exsig'].diff().shift(1) == 1)
df.loc[entsig_mask, 'ent'] = 1
df.loc[exsig_mask, 'ex'] = 1
df['in'] = df['ent'].shift(1).cumsum().subtract(df['ex'].cumsum(), fill_value=0)
return df
If we can make this assumption, then ent_in_ex is significantly faster:
In [5]: %timeit orig(df)
10 loops, best of 3: 185 ms per loop
In [6]: %timeit ent_in_ex(df)
100 loops, best of 3: 2.23 ms per loop
In [95]: orig(df).equals(ent_in_ex(df))
Out[95]: True
where orig is the original code:
def orig(df):
for i in df.index:
df['ent'][(df.entsig.shift(1)==1) & (df['ent'].shift(1) == 0) & (df['in'].shift(1) == 0)]=1
df['ex'][(df.exsig.shift(1)==1) & (df['in'].shift(1)==1)]=1
df['in'][(df.ent.shift(1)==1) | ((df['in'].shift(1)==1) & (df['ex']==0))]=1
for j in df.index:
df['ent'][df['in'] == 1]=0
df['in'][df['ex']==1]=0
df['ex'][df['ex'].shift(1)==1]=0
return df