0

I am trying to build a predictive model for a bank that needs to decide who will be approved for a loan based on the applicant's demographic and socio-economic profiles. The dataset has columns that detail an applicant's age, employment, credit history, etc. (a mix of objects and integers).

Below is what I've done up until the Valuerror:

import pandas as pd
import matplotlib.pyplot as plt
import numpy as np

from sklearn.model_selection import train_test_split
from sklearn.metrics import confusion_matrix
from sklearn.metrics import accuracy_score

import seaborn as sns

import warnings
warnings.filterwarnings("ignore")

credit = pd.read_csv('german_credit_risk.csv')

credit.head(10)

Image of output

credit.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 1000 entries, 0 to 999
Data columns (total 21 columns):
 #   Column             Non-Null Count  Dtype 
---  ------             --------------  ----- 
 0   Checking.Account   1000 non-null   object
 1   Duration           1000 non-null   int64 
 2   Credit.Hist        1000 non-null   object
 3   Purpose            1000 non-null   object
 4   Credit.Amount      1000 non-null   int64 
 5   Saving.Account     1000 non-null   object
 6   Employmnet         1000 non-null   object
 7   Disposable.Income  1000 non-null   int64 
 8   Personal.Status    1000 non-null   object
 9   Other.Debtors      1000 non-null   object
 10  Present.Residence  1000 non-null   int64 
 11  Property           1000 non-null   object
 12  Age                1000 non-null   int64 
 13  Other.Loans        1000 non-null   object
 14  Housing            1000 non-null   object
 15  Existing.Credits   1000 non-null   int64 
 16  Job                1000 non-null   object
 17  Number.Liable      1000 non-null   int64 
 18  Telephone          1000 non-null   object
 19  Foreign.Worker     1000 non-null   object
 20  Cost.Matrix        1000 non-null   int64 
dtypes: int64(8), object(13)
memory usage: 164.2+ KB

credit.describe(include='all')

image of output

_ = sns.heatmap(credit.corr())

image of output

col_names = ['Checking.Account', 'Duration', 'Credit.Hist', 'Purpose', 'Credit.Amount', 'Saving.Account', 'Employment', 'Disposable.Income', 'Personal.Status', 'Other.Debtors', 'Present.Residence', 'Property', 'Age', 'Other.Loans', 'Housing', 'Existing.Credits', 'Job', 'Number.Liable', 'Telephone', 'Foreign.Worker', 'Cost.Matrix']

credit.columns = col_names

cols = credit.columns

prediction_col = 'Credit.Hist'
feature_cols = [c for c in cols if c !=prediction_col]

x = credit[feature_cols].values
y = credit[prediction_col].values

x_train, x_test, y_train, y_test = train_test_split(x, y, test_size=0.2, random_state=41)

x_train, x_test, y_train, y_test = train_test_split(x, y, test_size=0.2, random_state=41)

from sklearn import tree

tree_model = tree.DecisionTreeClassifier()
tree_model.fit(x_train, y_train)

ValueError below:

ValueError                                Traceback (most recent call last)
/tmp/ipykernel_30064/311858196.py in <module>
      2 
      3 tree_model = tree.DecisionTreeClassifier()
----> 4 tree_model.fit(x_train, y_train)

~/anaconda3/lib/python3.7/site-packages/sklearn/tree/_classes.py in fit(self, X, y, sample_weight, check_input, X_idx_sorted)
    940             sample_weight=sample_weight,
    941             check_input=check_input,
--> 942             X_idx_sorted=X_idx_sorted,
    943         )
    944         return self

~/anaconda3/lib/python3.7/site-packages/sklearn/tree/_classes.py in fit(self, X, y, sample_weight, check_input, X_idx_sorted)
    164             check_y_params = dict(ensure_2d=False, dtype=None)
    165             X, y = self._validate_data(
--> 166                 X, y, validate_separately=(check_X_params, check_y_params)
    167             )
    168             if issparse(X):

~/anaconda3/lib/python3.7/site-packages/sklearn/base.py in _validate_data(self, X, y, reset, validate_separately, **check_params)
    571                 # :(
    572                 check_X_params, check_y_params = validate_separately
--> 573                 X = check_array(X, **check_X_params)
    574                 y = check_array(y, **check_y_params)
    575             else:

~/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py in check_array(array, accept_sparse, accept_large_sparse, dtype, order, copy, force_all_finite, ensure_2d, allow_nd, ensure_min_samples, ensure_min_features, estimator)
    736                     array = array.astype(dtype, casting="unsafe", copy=False)
    737                 else:
--> 738                     array = np.asarray(array, order=order, dtype=dtype)
    739             except ComplexWarning as complex_warning:
    740                 raise ValueError(

ValueError: could not convert string to float: 'A12'

I think I understand why the error came up, but I'm still lost. It's because I'm dealing with both integers and objects, correct? Where I'm lost is 'A12'. A12 is varaiable that comes up in multiple columns. So which one? And why that particular one as opposed to A11 or A13 or one of the other variables?

pete-ar
  • 1
  • 1
    sklearn models cannot take text as input. You need to encode text input to numeric in order to `fit` a model to it. LabelEncoder, OneHotEncoder, manual mapping, etc – G. Anderson Feb 04 '22 at 16:36
  • @G.Anderson Thank you for clarifying. So would the following be sufficient: #Import LabelEncoder. `from sklearn.preprocessing import LabelEncoder` `le = LabelEncoder()` #Create df to populate with encoded values. `encoded_obj_df = pd.DataFrame(columns=obj_cols)` #Apply the transformer. `for col in obj_cols: encoded_obj_df[col] = le.fit_transform(df[col])` `encoded_obj_df.head(10)` #check the results. This is something I'm doing prior to establishing the prediction column? – pete-ar Feb 04 '22 at 16:48
  • Comments are terrible for code formatting, so it's hard to say, but something like that should help yes. Just be sure you `fit` the transformer to your training data, and then use it to `transform` your training and test data, to prevent data leakage – G. Anderson Feb 04 '22 at 17:04

0 Answers0