# How to read data from external files into R and save them as a time series (ts) object?

# Simple but tedious: typing data as a vector (only useful for very short series):

my.data = c(12,31,22,24,30)

# Converting it to a 'ts' object (Suppose these are QUARTERLY data beginning at the second quarter of 2006):

dat2 = ts(my.data, start=c(2006,2), frequency=4)

# Check the created object:

print(dat2)  
# print the observed data values

time(dat2) 
# print the times at which the measurements were taken
########################################################

# What if the data are in an external data file, with the data values separated by spaces, tabs, or line breaks?

# Suppose the file were called 'file1.txt':

# Note:  before running this code, go to 
# File -> Change dir
# in R, and navigate to the directory on your computer in which 'file1.txt' is stored.
# On a Mac, the correct menus to go to in order to change directory are:
# Misc --> Change Working Directory 
# or in RStudio on a Mac:
# Session --> Set Working Directory --> Choose Directory

dat2 = ts(scan('file1.txt'), start=c(2006,2), frequency=4)

# What if there were several (say, 3) lines at the top of the data file that were just 
# descriptions of the data, not real data values, which we want to skip over?

# Use the 'skip' argument:

dat2 = ts(scan('file1.txt', skip=3), start=c(2006,2), frequency=4)

# What if the data are in a multicolumn table, with one column containing the time values 
# and the other column containing the data values, like this:

Time		Response
2006.Qtr2	12
2006.Qtr3	31
2006.Qtr4	22
2007.Qtr1	24
2007.Qtr2	30

# Suppose the file were called 'file2.txt':

# Note:  On WINDOWS, before running the following code, go to 
# File -> Change dir
# in R, and navigate to the directory on your computer in which 'file2.txt' is stored.
# On a Mac, see above for the instructions for changing directory.

dat3 = read.table('file2.txt', header=T)
# this is if the first line of the data file is a column with the variable names

dat3 = read.table('file2.txt', header=F)
# this is if the data start on the first line, with no header

# Then converting it to a 'ts' object:

Yvec = dat3[,2]  
# picks out 2nd column of dat3

dat3.ts = ts(Yvec, start=c(2006,2), frequency=4)

# What if the data are in an Excel file?
# First save the Excel file as a .csv file.

# Note:  before running this code, go to 
# File -> Change dir
# in R, and navigate to the directory on your computer in which 'file3.csv' is stored.

dat4 = read.csv('file3.csv', header=T)

# Then convert to data vector to a 'ts' object as shown above.

# Please read the section labeled "Reading time series data" on the web page that I link to 
# on the link labeled "Examples for reading in Time Series and creating 'ts' objects in R"
# on our course web page!!