How can I add a string in nparray

Question

Here's my code:

age=request.user.allusers.age
sex=request.user.allusers.sex
index_array = np.array(states)
index_array = [val-1 for val in index_array]
mask_array = np.zeros(193,dtype=float)
mask_array[index_array] = 1
mask_array=np.append(sex,mask_array)
mask_array=np.append(age,mask_array)
prob_array = clf.predict_proba([mask_array])

In the code sex is a string. So when compiling the value error is showing up. How can I append sex in the numpy array.

Here's the error:

ValueError: could not convert string to float: 'Female'

Show where the error occurs. What's the dtype of the arrray? Strings can only be added to string dtype arrays or structured arrays (compound dtype). What does `clf` accept? — hpaulj, Mar 24 '18 at 15:24
That's a bad duplicate. We don't know that the OP needs a structured array. — hpaulj, Mar 24 '18 at 15:49

Jai · Accepted Answer · 2018-03-24T09:37:23.863

Simply insert the values inside the list

Give list as an argument to numpy array

import numpy as np

age=23
sex="Female"

info_list = [age, sex]
print("This is list", info_list)

numpy_info_list = np.array(info_list)
print("This is numpy array", numpy_info_list)

Output:

This is list [23, 'Female']
This is numpy array ['23' 'Female']

Usually this is not good for predicting to give the data in the raw form
Instead you can convert Female to ===> 0 and Male to ====> 1
We should always try to convert the values to numbers
Sex is a type of categorical data which can take numerical values
One of the solution is to use scikitlearn label encoder for encoding the features if you have too may values for a particular features
I know its stupid to use a label encoder to encode the features but it's a solution

Here's an example... This is just an example related to sex

le = preprocessing.LabelEncoder()
>>> le.fit(["male", "female"])
LabelEncoder()
>>> list(le.classes_)
['male', 'female']
>>> le.transform(["male", "male", "female"]) 
array([1, 1, 2]...)

Other solution can be to use dictionary:

age=23
sex="female"
sex_map = {"male": 1, "female": 2}

info_list = [age, sex_map[sex]]
print("This is list", info_list)

Output:
```
This is list [23, 2]
```

How can I add a string in nparray

1 Answers1