Homework #1: Firebase and JSON

INF 551
Homework #1: Firebase and JSON

100 points
In this homework, we consider the LA restaurants health inspection data set: restaurants.csv, in the CSV
format. We are only concerned with two columns in the data set: serial_number, facility_name and
score.
1. [60 points] Write a Python script (with REST requests embedded) called “load.py”. The script will do
two things:
o Convert the data (only the needed three columns) into the JSON format and load the
dataset into Firebase. You may need Python “requests” package as shown in class.
o Create an inverted index for the facility_name column. The index stores, for each unique
word in the name (you can assume that words are delimited by white spaces and
punctuation characters), the serial_number of restaurant whose name contains the
word.
For example, the name of the first restaurant (serial number = DAJ00E07B) has 3 unique
words: habitat, coffee, and shop. You should lower case the words in the index. The
index looks like the following:
{"index": {
"habitat": [DAJ00E07B, …],
“coffee”: [DAJ00E07B, …],
“shop”: [DAJ00E07B, …],
…
}
Execution format:
o python load.py restaurants.csv
• [40 points] Write a Python script called “search.py”. The script takes a list of keywords and
returns names and scores of restaurants whose name contains one or more keywords in the list.
The search needs to be executed using the data stored in your Firebase database and use the
above index. Note that the search is NOT case-sensitive. For example,
o python search.py “coffee shop”
should return the restaurants whose name contains “coffee” or “shop” or both.
INF 551 – Fall 2019
Submissions: Name your 2 scripts as below and submit to Blackboard by the due time. DO NOT place
them in a folder or zip file.
• <FirstName_<LastName_load.py
• <FirstName_<LastName_search.py
Note: Please use Python 3.6 for your homework. To install Python 3.6 on EC2, execute this:
sudo yum install python36 python36-pip
However, please do not remove Python2 from EC2, which may be needed for Spark.
To execute the new Python, type: python3, instead of python.
Note that the new usage of pip in Python 3:
sudo python3 -m pip install <package-name
The Python packages allowed in this homework are: requests, json, pandas, and numpy.
Output formats:
For <FirstName_<LastName_load.py, print the response in json format after using requests.put to
upload your inverted index to the database. E.g.
For <FirstName_<LastName_search.py, print the response in json format. E.g.