UDACITY DISASTER RESPoNSE PIPELINES : 

The Project contains two datasets , categories and messages , The project classifies messages on the basis of keywords and assigns them to the respective category. THe first part of the project contains Cleaning of data using NLP pipelines. ETLpreparation The messages are tokenized , lematized using tokenize text function. The CAtegories columns are extracted from the data and new columns are assigned on the basis of respective categories. The new Data set is formed by merging categories data set with messages The libraries used in the ETL pipeline are

1. pandas 2.sql alchemy engine

The second part of the project contains creating a machine learning model that classifies the messages with thier respective categories pipeline is used so the CountVectorization, FeatureExtraction and Tfidf can be done parallely. GridSeacrh CV is used so that best parameters can be choosen to train the dataset. Decision tree classifier and random forest classifier are used in training the model in which random forest classifier gave better values of recall precision, accuracy and f1 score and was included in train_Classifier.py ML-prepration File contains both the models but classification is done on the basis of Random forest classifier. The libraries included in ML pipeline are :

sqlalchemy import create_engine numpy as np sklearn.pipeline import Pipeline sklearn.feature_extraction.text import CountVectorizer, TfidfTransformersklearn.model_selection import train_test_split, GridSearchCV sklearn.ensemble import RandomForestClassif sklearn.pipeline import FeatureUnion sklearn.multioutput import MultiOutputClassifier sklearn.metrics import accuracy_score, precision_score, recall_score, f1_score, make_scorer sklearn.model_selection import GridSearchCV sklearn.tree import DecisionTreeClassifier import warnings import re import pickle import nltk nltk.corpus import stopwords nltk.tokenize import word_tokenize nltk.stem.porter import PorterStemmer nltk.stem.wordnet import WordNetLemmatizer sklearn.metrics import classification_report

                                     RUNNING THE APP :
                                     
       1 . Clone the repository using  git clone https://github.com/yashmanuraj2/disatster-response-pipeline.git
       2. In cd /data run process_data.py
       3. In cd/ models run train_classifier.py
       4. IN app run app.py
       5. type env | grep WORK in new terminal and add ID - domain as the domain name
         https://view6914b2f4-3001.udacity-student-workspaces.com
       id - view6914b2f4-3001
       domain - udacity-student-workspaces.com/ 
       App is running on port 3001