Model-1.py

# -*- coding: utf-8 -*-
"""Working_Model.ipynb

Automatically generated by Colaboratory.

Original file is located at
    https://colab.research.google.com/drive/15vMWS3F0cKUYSGMFpdKrZkNjuDhfApRm
"""

from google.colab import drive
drive.mount('/content/drive')

import pandas as pd
import numpy as np


import matplotlib.pyplot as plt
import seaborn as sns


from sklearn.preprocessing import *
from sklearn.model_selection import *
from sklearn.metrics import *

from xgboost import XGBClassifier
from sklearn.ensemble import GradientBoostingClassifier
from lightgbm import LGBMClassifier
from sklearn.ensemble import VotingClassifier

train = pd.read_csv("/content/drive/My Drive/train.csv", encoding="utf-8")
test = pd.read_csv("/content/drive/My Drive/test.csv", encoding="utf-8")

train.dtypes

def data_info(data, data_types, dataframe_name):
    print(" Information of ",dataframe_name,": Rows = ",data.shape[0],"| Columns = ",data.shape[1],"\n")
    data.info()
    print("\n")
    for VARIABLE in data_types:
        data_type = data.select_dtypes(include=[VARIABLE]).dtypes
        if len(data_type) > 0:
            print(str(len(data_type))+" "+VARIABLE+" Features\n"+str(data_type)+"\n"  )        


data_types  = ["float32","float64","int32","int64","object","category","datetime64[ns]"]
data_info(train, data_types, "train")

def display_head_tail(data, head_size=5, tail_size=5):
  print("The head and tail of the data is given as follows:")
  display(data.head(head_size).append(data.tail(tail_size)))

display_head_tail(train, 3, 3)

def remove_duplicates(data):
  print("Before Removing Duplicate Data: ", data.shape[0])
  data.drop_duplicates(keep='first', inplace=True)
  print("After Removing Duplicate Data: ", data.shape[0])
remove_duplicates(train)

def handle_missing_values(data, fill_value, fill_types, columns, dataframe_name):
  print("Missing Values BEFORE REMOVAL in ", dataframe_name)
  display(data.isna().sum())
  for column in columns :
        
      # Fill Missing Values with Specific Value :
      if "Value_Fill" in fill_types :
          data[ column ] = data[ column ].fillna(fill_value)
#            print("Value_Fill")

        # Fill Missing Values with Forward Fill  (Previous Row Value as Current Row in Table) :
      if "Forward_Fill" in fill_types :
          data[ column ] = data[ column ].ffill(axis = 0)
#            print("Forward_Fill")

        # Fill Missing Values with Backward Fill (Next Row Value as Current Row in Table) :
      if "Backward_Fill" in fill_types :
          data[ column ] = data[ column ].bfill(axis = 0)
#             print("Backward_Fill")
    
  print("Missing Values AFTER REMOVAL in ",dataframe_name," data")
  display(data.isnull().sum())
    
  return data

fill_types = [ "Forward_Fill"]
fill_value = train["Number_Weeks_Used"].median()
train = handle_missing_values(train, fill_value, fill_types, ["Number_Weeks_Used"],"train")

def unique_values(data):
  for column in data.columns:
    print("Number of unique values are: ", str(data[column].nunique()))
    print("Actual Unique Values in "+column+" Column are : "+
          str(data[column].sort_values(ascending=True,na_position='last').unique() ))
    print("Value Counts :")
    print(data[column].value_counts())
    print("")
unique_values(train)

def heatmap(df):
  plt.figure(figsize=(10,10))
  corr = df.corr()
  return sns.heatmap(corr, annot=True, cmap='viridis')

def cntplt(df, feature):
  plt.figure(figsize=(10, 10))
  return sns.countplot(data=df, x=df[feature])

heatmap(train)
cntplt(train, 'Crop_Damage')

sns.catplot(x="Crop_Damage", y="Season", hue="Crop_Damage", kind="bar", data=train);

X = train.drop(labels=['Crop_Damage'], axis=1)
y = train['Crop_Damage']

from sklearn.model_selection import train_test_split
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3)

y_test.value_counts

X_train_data = X_train.iloc[:, 1:9]
X_test_data = X_test.iloc[:, 1:9]
X_train_data

lgbm = LGBMClassifier()
lgbm_pred = lgbm.fit(X_train_data, y_train)
y_pred = lgbm_pred.predict(X_test_data)
print(accuracy_score(y_pred, y_test))

test2 = test.iloc[:, 1:9]
test_pred = lgbm_pred.predict(test2)
test2['Crop_Damage'] = test_pred
test2['ID'] = test['ID']
test2

output=pd.DataFrame(data={"ID":test2["ID"],"Crop_Damage":test2["Crop_Damage"]}).to_csv("Sol.csv", index=False)
from google.colab import files
files.download('Sol.csv')

output

output=pd.DataFrame(data={"ID":["ID"],"Crop_Damage":y_pred}).to_csv("Sample.csv", index=False)
from google.colab import files
files.download('Sample.csv')