TY.B.Sc[COMPUTER SCIENCE] Roll No.

1952034

Certificate

VIDYA PRASARAK MANDAL’S

R.Z. SHAH COLLEGE OF ARTS, SCIENCE &
COMMERCE
Mithagar Road,Mulund(E) - 400 081

This is to certify that Mr./Miss Ritesh Shukla Roll No.1952034 has

completed his journal in the subject of DATA SCIENCE under the program
of BSC.Computer Science Semester SEM-VI University of Mumbai during
the year 2020-2021
He has completed the prescribed practical satisfactorily.

Examination Seat No.1952014

Date: _________________

In charge Head of the Department

Sr. Title Date Signature

No

Sub:DATA SCIENCE
TY.B.Sc[COMPUTER SCIENCE] Roll No. 1952034

1 Practical of Data 7/02/2020

collection, Data curation and
management for Unstructured data
(NoSQL)

2 Practical of Data collection, Data 31/01/2020

curation and management for Large-
scale Data system (such as MongoDB)

7/02/2020
3 Practical of Principal Component
Analysis

4 Practical of Clustering 24/01/2020

5 Practical of Time-series forecasting 17/01/2020

6 Practical of Simple/Multiple Linear 24/01/2020

Regression
7/02/2020
7 Practical of Logistics Regression
14/02/2020
8 Practical of Hypothesis testing
21/02/2020
9 Practical of Analysis of Variance
28/02/2020
10 Practical of Decision Tree

INDEX

Practical No. 1 Date: 7/02/2020

Aim: Practical of Data collection, Data curation and management for Unstructured data
(NoSQL)
Sub:DATA SCIENCE
TY.B.Sc[COMPUTER SCIENCE] Roll No. 1952034

Couchdb database ---

rscript Install

couchdb first

Rscript code

install.packages('sofa')

#devtools::install_github("rop

ensci/sofa") library('sofa')

#create

connection

object x<-

Cushion$n

ew()

#to check whether

object created

x$ping()

#create

database ty

db_create(x,dbna

Sub:DATA SCIENCE
TY.B.Sc[COMPUTER SCIENCE] Roll No. 1952034

me = 'ty')

db_list(x)

#create json doc

doc1<-'{"rollno":"01","name":"ABC","GRADE":"A"}'

doc_create(x,doc1,dbname

= "ty",docid = "a_1")

doc2<-'{"rollno":"02","name":"

PQR","GRADE":"A"}'

doc_create(x,doc2,dbname = "ty",docid = "a_2")

doc3<-'{"rollno":"03","name":"xyz","GRADE":"B","REMARK":"PASS"}'

doc_create(x,doc3,dbname = "ty",docid = "a_3")

#CHANGES FEED

db_changes(x,"ty")

#search for id > null

so all docs will

display

db_query(x,dbname

= "ty",

selector = list('_id'=list('$gt'=NULL)))$docs

Sub:DATA SCIENCE
TY.B.Sc[COMPUTER SCIENCE] Roll No. 1952034

#search for students with grade is A

db_query(x,dbname = "ty",selector = list(GRADE="A"))$docs

#search for students with remark =pass

db_query(x,dbname = "ty",selector = list(REMARK="PASS"))$docs

#return only certain fields where rollno>2

db_query(x,dbname = "ty",selector =
list(rollno=list('$gt'='02')),fields=c("name","GRADE"))$docs

#convert the result of a query

into a data frame using jsonlite

library("jsonlite")

res<-db_query(x,dbname = "ty",selector =
list('_id'=list('$gt'=NULL)),fields=c("name","rollno","GRADE","REMARK"),as="json")

Sub:DATA SCIENCE
TY.B.Sc[COMPUTER SCIENCE] Roll No. 1952034

#display json doc fromJSON(res)$docs

#doc_delete(cushion,d

bname,docid)

doc_delete(x,dbname =

"ty",docid = "a_2")

doc_get(x,dbname =

"ty",docid = "a_2")

doc2<-'{"name":"Sdrink","beer":"TEST","note":"yummy","note2":"yay"}'

doc_update(x,dbname = "ty",doc=doc2,docid="a_3",rev = "3-

b1fb56db955b142c6efd3b3c52fe9e1b")

doc3<-'{"rollno":"0

1",

"name":"UZMA",

"GRADE":"A"}'

doc_update(x,dbname = "ty",doc=doc3,docid = "a_1",rev = "1-

be7c98bddf8ea7c46f4f401ff387593d")

Sub:DATA SCIENCE
TY.B.Sc[COMPUTER SCIENCE] Roll No. 1952034

Output:

Sub:DATA SCIENCE
TY.B.Sc[COMPUTER SCIENCE] Roll No. 1952034

Sub:DATA SCIENCE
TY.B.Sc[COMPUTER SCIENCE] Roll No. 1952034

Sub:DATA SCIENCE
TY.B.Sc[COMPUTER SCIENCE] Roll No. 1952034

Sub:DATA SCIENCE
TY.B.Sc[COMPUTER SCIENCE] Roll No. 1952034

Sub:DATA SCIENCE
TY.B.Sc[COMPUTER SCIENCE] Roll No. 1952034

Sub:DATA SCIENCE
TY.B.Sc[COMPUTER SCIENCE] Roll No. 1952034

Sub:DATA SCIENCE
TY.B.Sc[COMPUTER SCIENCE] Roll No. 1952034

Practical No. 2 Date: 31/01/2020

Aim: Practical of Data collection, Data curation and management for
Large-scale Data system (such as MongoDB)

 MongoDB Create database

 MongoDB Drop Database
 MongoDB Create collection
 MongoDB Drop collection
 MongoDB Insert Document
 MongoDB Query Document
 MongoDB Update Document
 Delete document in MongoDB
 MongoDB Projection
 limit() and skip() method in MongoDB
 Sorting of Documents in MongoDB
 MongoDB Indexing

Starting server with mongo or mongodb

C:\>mongo
>db
Test

 Create Database in MongoDB

Once you are in the MongoDB shell, create the database in MongoDB
by typing this command:
use database_name

For example: create a database “tycs”:

> use tycs

switched to db tycs

Sub:DATA SCIENCE
TY.B.Sc[COMPUTER SCIENCE] Roll No. 1952034

create a collection user and insert a document in it.

> db.user.insert({name: "Asif", age: 20})

O/P: WriteResult({ "nInserted" : 1 })

>show dbs
admin 0.000GB
config 0.000GB
local 0.000GB
tycs 0.000GB

 MongoDB Drop Database

The syntax to drop a Database is:

>db.dropDatabase()

O/P:
{ "dropped" : "Testdb", "ok" : 1 }
MongoDB Enterprise > show dbs
admin 0.000GB
config 0.000GB
local 0.000GB
tycs 0.000GB

Sub:DATA SCIENCE
TY.B.Sc[COMPUTER SCIENCE] Roll No. 1952034

O/P:

 Create Collection in MongoDB

Method 1: Creating the Collection in MongoDB on the fly

MongoDB Enterprise > use tycs

switched to db tycs

MongoDB Enterprise > db.tycs.insert({name:"Asif

khan",age:21,website:"www.google.com"})
O/P:
WriteResult({ "nInserted" : 1 })

Syntax: db.collection_name.find()
MongoDB Enterprise > db.tycs.find()

o/p:
{ "_id" : ObjectId("5e410808e3755b1e06a63d1d"), "name" : "Asif khan",
"age" : 21, "website" : "www.google.com" }

show collections

MongoDB Enterprise > show collections

Sub:DATA SCIENCE
TY.B.Sc[COMPUTER SCIENCE] Roll No. 1952034

O/P:
tycs
user

 Drop collection in MongoDB

SYNTAX:
db.collection_name.drop()

MongoDB Enterprise > use students

switched to db students

MongoDB Enterprise > show collections

students
teachers
tycs
user

MongoDB Enterprise > db.user.drop()

true
MongoDB Enterprise > show collections
students
teacher
tycs
MongoDB Insert Document

Syntax to insert a document into the collection:

db.collection_name.insert()

Sub:DATA SCIENCE
TY.B.Sc[COMPUTER SCIENCE] Roll No. 1952034

> db.tycs.insert(
... {
... name: "ASIF",
... age: 20,
... email: "asif@gmail.com",
... course: [ { name: "MongoDB", duration: 7 }, { name: "Java",
duration: 30 } ]
... }
... )

O/P:
WriteResult({ "nInserted" : 1 })

Verification:
Syntax:
db.collection_name.find()

> db.tycs.find()
{ "_id" : ObjectId("5c2d37734fa204bd77e7fc1c"), "name" : "ASIF",
"age" : 20, "email" : "asif@gmail.com", "course" : [ { "name" :
"MongoDB", "duration" : 7 }, { "name" : "Java", "duration" : 30 } ] }

MongoDB Example: Insert Multiple Documents in collection

MongoDB Enterprise > var beginners=

... [
... "studentID":1001,
... "studentName":"Asif",
... "age":20
... },
... ]

 MongoDB Query Document using find() method

Querying all the documents in JSON format

MongoDB Enterprise > db.students.find().pretty()

Sub:DATA SCIENCE
TY.B.Sc[COMPUTER SCIENCE] Roll No. 1952034

{
"_id" : ObjectId("5e410f3fe3755b1e06a63d1e"),
"studentID" : 1001,
"studentName" : "Asif",
"age" : 20
}

 Query Document based on the criteria

> db.students.find({StudentName : "Asif"}).pretty()

{
"_id" : ObjectId("5c281c90c23e08d1515fd9cc"),
"StudentId" : 1001,
"StudentName" : "Asif",
"age" : 20
}

 Updating Document using update() method

Syntax:

db.collection_name.update(criteria, update_data)

> use tycs

switched to db tycs
> show collections
beginnersbook
students
tycs
> db.createCollection("got")
{ "ok" : 1 }
> var abc = [
... {
... "_id" : ObjectId("59bd2e73ce524b733f14dd65"),
... "name" : "Asif",
... "age" : 20
... },
... ];

Sub:DATA SCIENCE
TY.B.Sc[COMPUTER SCIENCE] Roll No. 1952034

> db.got.find().pretty()
{
"_id" : ObjectId("59bd2e73ce524b733f14dd65"),
"name" : "steve",
"age" : 20
}

To update multiple documents with the update() method:

db.got.update({"name":"Jon Snow"},
{$set:{"name":"Kit Harington"}},{multi:true})

Updating Document using save() method

Syntax:

db.collection_name.save( {_id:ObjectId(), new_document} )

To get the _id of a document, you can either type this command:

db.got.find().pretty()

> db.got.find({"name": "Asif"}).pretty()

{
"_id" : ObjectId("59bd2e73ce524b733f14dd65"),
"name" : "Asif",
"age" : 20
}
> db.got.find().pretty()
{
"_id" : ObjectId("59bd2e73ce524b733f14dd65"),
"name" : "Steve",
"age" : 20
}

 MongoDB Delete Document from a Collection

Syntax of remove() method:

Sub:DATA SCIENCE
TY.B.Sc[COMPUTER SCIENCE] Roll No. 1952034

db.collection_name.remove(delete_criteria)

Delete Document using remove() method

> db.students.find().pretty()
{
"_id" : ObjectId("59bcecc7668dcce02aaa6fed"),
"StudentId" : 1001,
"StudentName" : "Steve",
"age" : 30
}

db.students.remove({"StudentId": 3333})
Output:

WriteResult({ "nRemoved" : 1 })
To verify whether the document is actually deleted. Type the following
command:

db.students.find().pretty()
It will list all the documents of students collection.

> use tycs

switched to db tycs
> db.students.find().pretty()
{
"_id" : ObjectId("5c281c90c23e08d1515fd9cc"),
"StudentId" : 1001,
"StudentName" : "Asif",
"age" : 20
}
{
"_id" : ObjectId("5c2d38934fa204bd77e7fc1d"),
"StudentId" : 1001,
"StudentName" : "Steve",
"age" : 30
}

Remove all Documents

db.collection_name.remove({})

Sub:DATA SCIENCE
TY.B.Sc[COMPUTER SCIENCE] Roll No. 1952034

 MongoDB Projection

Syntax:

db.collection_name.find({},{field_key:1 or 0})

> db.students.find().pretty()
{
"_id" : ObjectId("5c281c90c23e08d1515fd9cc"),
"StudentId" : 1001,
"StudentName" : "Steve",
"age" : 20
}
> db.students.find({}, {"_id": 0, "StudentId" : 1})
{ "StudentId" : 1001 }
{ "StudentId" : 1002 }

> db.students.find({}, {"_id": 0, "StudentName" : 0, "age" : 0})

{ "StudentId" : 1001 }
{ "StudentId" : 1002 }
 MongoDB – limit( ) and skip( ) method

The limit() method in MongoDB

Syntax:

db.collection_name.find().limit(number_of_documents)
db.studentdata.find({student_id : {$gt:2002}}).pretty()
db.studentdata.find({student_id : {$gt:2002}}).limit(1).pretty()

Sub:DATA SCIENCE
TY.B.Sc[COMPUTER SCIENCE] Roll No. 1952034

MongoDB Skip() Method

db.studentdata.find({student_id : {$gt:2002}}).limit(1).skip(1).pretty()

 MongoDB sort() method

Sorting Documents using sort() method

Syntax of sort() method:

db.collecttion_name.find().sort({field_key:1 or -1})
1 is for ascending order and -1 is for descending order. The default
value is 1.

For example: collection studentdata contains following documents:

> db.studentdata.find().pretty()

{
"_id" : ObjectId("59bf63380be1d7770c3982af"),
"student_name" : "Steve",
"student_id" : 1001,
"student_age" :1002
}

Let’s display the student_id of all the documents in descending order:

> db.studentdata.find({}, {"student_id": 1, _id:0}).sort({"student_id": -1})
{ "student_id" : 1001 }
{ "student_id" : 1002 }

To display the student_id field of all the students in ascending order:

> db.studentdata.find({}, {"student_id": 1, _id:0}).sort({"student_id": 1})

{ "student_id" : 1001 }
{ "student_id" : 1002 }

Sub:DATA SCIENCE
TY.B.Sc[COMPUTER SCIENCE] Roll No. 1952034

 MongoDB Indexing with Example

How to create index in MongoDB

db.collection_name.createIndex({field_name: 1 or -1})
The value 1 is for ascending order and -1 is for descending order.

Let’s create the index on student_name field in ascending order:

db.studentdata.createIndex({student_name: 1})

Output:

{
"createdCollectionAutomatically" : false,
"numIndexesBefore" : 1,
"numIndexesAfter" : 2,
"ok" : 1
}

 MongoDB – Finding the indexes in a collection

db.collection_name.getIndexes()
> db.studentdata.getIndexes()
[

Sub:DATA SCIENCE
 TY.B.Sc[COMPUTER SCIENCE] Roll No. 1952034

{
"v" : 2,
"key" : {
"_id" : 1
},
"name" : "_id_",
"ns" : "test.studentdata"
},
]

Practical No. 3 Date :7/02/2020

Aim: Practical of Principal Component Analysis

data_iris<-iris[1:4]
Cov_data<-cov(data_iris)
Eigen_data<-eigen(Cov_data)
PCA_data<-princomp(data_iris,cor="False")
Eigen_data$values
PCA_data$sdev^2

Sub:DATA SCIENCE
TY.B.Sc[COMPUTER SCIENCE] Roll No. 1952034

PCA_data$loadings[,1:4]
Eigen_data$vectors
summary(PCA_data)
biplot(PCA_data)
screeplot(PCA_data,type="lines")
model2=PCA_data$loadings[,1]
model2_scores<-as.matrix(data_iris)%*%model2
library(class)
install.packages("e1071")
library(e1071)

mod1<-naiveBayes(iris[,1:4],iris[,5])
mod2<-naiveBayes(model2_scores,iris[,5])
table(predict(mod1,iris[,1:4]),iris[,5])
table(predict(mod2,model2_scores),iris[,5])

Output:

Sub:DATA SCIENCE
TY.B.Sc[COMPUTER SCIENCE] Roll No. 1952034

Sub:DATA SCIENCE
TY.B.Sc[COMPUTER SCIENCE] Roll No. 1952034

Sub:DATA SCIENCE
TY.B.Sc[COMPUTER SCIENCE] Roll No. 1952034

Practical No. 4 Date :24/01/2020

Aim: Practical of Clustering

"K-means Clustering "

data(iris)
names(iris)
new_data<-subset(iris,select = c(-Species))
new_data
cl<-kmeans(new_data,3)
cl
data<-new_data
wss<-sapply(1:15,function(k){kmeans(data,k)$tot.withinss})
wss
plot(1:15,wss,type="b",pch=19,frame=FALSE,xlab ="Number of
clusters K",ylab = "Total within-clusters sums of squares")
library(cluster)
clusplot(new_data,cl$cluster,color=TRUE,shade=TRUE,labels=2,lines=
0)
cl$cluster
cl$centers
"agglomarative clustering "
clusters<-hclust(dist(iris[,3:4]))
plot(clusters)
clusterCut<-cutree(clusters,3)
table(clusterCut,iris$Species)

Sub:DATA SCIENCE
TY.B.Sc[COMPUTER SCIENCE] Roll No. 1952034

Output:

Sub:DATA SCIENCE
TY.B.Sc[COMPUTER SCIENCE] Roll No. 1952034

Sub:DATA SCIENCE
TY.B.Sc[COMPUTER SCIENCE] Roll No. 1952034

Sub:DATA SCIENCE
TY.B.Sc[COMPUTER SCIENCE] Roll No. 1952034

Sub:DATA SCIENCE
 TY.B.Sc[COMPUTER SCIENCE] Roll No. 1952034

Practical No. 5 Date: 17/01/2020

Aim: Practical of Time-series forecasting

#consider the inbuilt data set Air Passengers

data("AirPassengers")
class(AirPassengers)
start(AirPassengers)
#to know the end of time series
end(AirPassengers)
frequency(AirPassengers)
#to know the mean median etc of the dataset
summary(AirPassengers)
#to plot the time series model
plot(AirPassengers)
abline(reg=lm(AirPassengers~time(AirPassengers)))
cycle(AirPassengers)
plot(aggregate(AirPassengers,FUN=mean))
boxplot(AirPassengers~cycle(AirPassengers))

Output:

Sub:DATA SCIENCE
TY.B.Sc[COMPUTER SCIENCE] Roll No. 1952034

Sub:DATA SCIENCE
TY.B.Sc[COMPUTER SCIENCE] Roll No. 1952034

Sub:DATA SCIENCE
 TY.B.Sc[COMPUTER SCIENCE] Roll No. 1952034

Practical No. 6 Date : 24/01/2020

Aim: Practical of Simple/Multiple Linear Regression

#consider some data set

height<-
c(102,117,105,141,135,115,138,144,137,100,131,119,115,121,113)
weight<-c(61,46,62,54,60,69,51,50,46,64,48,56,64,48,59)
#lm is for Linear Regression
student<-lm(weight~height)
student
#to predict use predict command
predict(student,data.frame(height=199),interval="confidence")
#to plot the data
plot(student)

Output:

Sub:DATA SCIENCE
TY.B.Sc[COMPUTER SCIENCE] Roll No. 1952034

Sub:DATA SCIENCE
TY.B.Sc[COMPUTER SCIENCE] Roll No. 1952034

Sub:DATA SCIENCE
 TY.B.Sc[COMPUTER SCIENCE] Roll No. 1952034

Practical No. 7 Date :7/02/2020

Aim: Practical of Logistics Regression

library(datasets)
ir_data<-iris
head(ir_data)
str(ir_data)
levels(ir_data$Species)
sum(is.na(ir_data))
ir_data<-ir_data[1:100,]
set.seed(100)
samp<-sample(1:100,80)
ir_test<-ir_data[samp,]
ir_ctrl<-ir_data[-samp,]
install.packages("ggplot2")
library(ggplot2)
library(ggplot2)
install.packages("GGally")
ggpairs(ir_test)
y<-ir_test$Species;
x<-ir_test$Sepal.Length
glfit<-glm(y~x,family='binomial')
summary(glfit)
newdata<-data.frame(x=ir_ctrl$Sepal.Length)
predicted_val<-predict(glfit,newdata,type="response")
prediction<-
data.frame(ir_ctrl$Sepal.Length,ir_ctrl$Species,predicted_val)
prediction

Sub:DATA SCIENCE
TY.B.Sc[COMPUTER SCIENCE] Roll No. 1952034

qplot(prediction[,1],round(prediction[,3]),col=prediction[,2],xlab='sepal.
Length',ylab='prediction using logistic Reg')

Output:

Sub:DATA SCIENCE
TY.B.Sc[COMPUTER SCIENCE] Roll No. 1952034

Sub:DATA SCIENCE
TY.B.Sc[COMPUTER SCIENCE] Roll No. 1952034

Sub:DATA SCIENCE
 TY.B.Sc[COMPUTER SCIENCE] Roll No. 1952034

Practical No. 8 Date :14/02/2020

Aim: Practical of Hypothesis testing

#Entering the data

>x=c(6.2,6.6,7.1,7.4,7.6,7.9,8,8.3,8.4,8.5,8.6,8.8,8.8,9.1,9.2,9.4,9.7,9.
9,10.2,10.4,10.8,11.3,11.9)

#one-sample Hypothesis test

>t.test(x-9,alternative = "two.sided",conf.level = 0.95)

#Two-sample hypothesis test

>x=c(481,421,421,422,425,427,431,434,437,439,446,447,448,454,46
3,465)
>y=c(429,430,430,431,36,437,440,441,445,446,447)

>test2<-t.test(x,y,alternative = "two.sided",mu=0,var.equal =
F,conf.level = 0.95)
>test2

Sub:DATA SCIENCE
TY.B.Sc[COMPUTER SCIENCE] Roll No. 1952034

Output:

 Interpretation of the result:

The p-value (0.2998) is greater than the significance level 5%(1-0.95)
So,we conclude that the null hypothesis that the population means
are equal is plausible.

Sub:DATA SCIENCE
TY.B.Sc[COMPUTER SCIENCE] Roll No. 1952034

Practical No.9 Date : 21/02/2020

Aim: Practical of Analysis of Variance

ftest<-read.csv(file.choose(),sep=",",header = T)
var.test(ftest$time_g1,ftest$time_g2,alternative = "two.sided")
//one way anova
names(data1)
summary(data1)
head(data1)
anv<-aov(formula=satindex~dept,data=data1)
summary(anv)
//two way anova
data2<-read.csv(file.choose(),sep=",",header=T)
names(data2)
summary(data2)
head(data2)
anv1<-aov(formula = satindex~dept+exp+dept*exp,data=data2)
summary(anv1)

Sub:DATA SCIENCE
TY.B.Sc[COMPUTER SCIENCE] Roll No. 1952034

Output:

Sub:DATA SCIENCE
TY.B.Sc[COMPUTER SCIENCE] Roll No. 1952034

Sub:DATA SCIENCE
TY.B.Sc[COMPUTER SCIENCE] Roll No. 1952034

Practical No.10 Date : 28/02/2020

Aim: Practical of Decision Tree

mydata<-data.frame(iris)
attach(mydata)
install.packages("rpart")
library(rpart)
model<-
rpart(Species~Sepal.Length+Sepal.Width+Petal.Length+Petal.Width,da
ta=mydata,method="class")
plot(model)
text(model,use.n=TRUE,all=TRUE,cex=0.8)
install.packages("tree")
library(tree)
model1<-
tree(Species~Sepal.Length+Sepal.Width+Petal.Length+Petal.Width,data
=mydata,method="class",split="gini")
plot(model1)
text(model,all=TRUE,cex=0.6)
install.packages("party")
library(party)
model2<-
ctree(Species~Sepal.Length+Sepal.Width+Petal.Length+Petal.Width,dat
a=mydata)
plot(model2)
library(tree)
mydata<-data.frame(iris)
attach(mydata)
model<-
tree(Species~Sepal.Length+Sepal.Width+Petal.Length+Petal.Width,data
=mydata,method="class",control=tree.control(nobs=150,mincut=10))

Sub:DATA SCIENCE
TY.B.Sc[COMPUTER SCIENCE] Roll No. 1952034

plot(model1)
text(model,all=TRUE,cex=0.6)
predict(model,iris)
model2<-
ctree(Species~Sepal.Length+Sepal.Width+Petal.Length+Petal.Width,dat
a=mydata,controls=ctree_control(maxdepth=2))
plot(model2)

Sub:DATA SCIENCE
TY.B.Sc[COMPUTER SCIENCE] Roll No. 1952034

Output:

Sub:DATA SCIENCE
TY.B.Sc[COMPUTER SCIENCE] Roll No. 1952034

Sub:DATA SCIENCE
TY.B.Sc[COMPUTER SCIENCE] Roll No. 1952034

Sub:DATA SCIENCE
TY.B.Sc[COMPUTER SCIENCE] Roll No. 1952034

Sub:DATA SCIENCE
TY.B.Sc[COMPUTER SCIENCE] Roll No. 1952034

Sub:DATA SCIENCE