Lesson on Mongo db @ University (Milan). From install packages on python, to test Pymongo on macOSX

mongoutils.sh

# ------ Prerequisites ------
# Start server
$ mongod 

# Start client
$ mongo

# Install packages on python 2.7 (required version of "pip" >= 10)
$ sudo pip install --ignore-installed pymongo
$ sudo pip install --ignore-installed numpy
$ sudo pip install --ignore-installed pandas

# Install pandas in python 3 (I use 3.6)
$ sudo python3 -m pip install pandas
$ sudo python3 -m pip show pandas



# If I query a collection that does not exist is created.
# To show dbs: 
$ show dbs

# The client contains a variable (db) that is initialized with the current db
$ use [dbname]

# I see the list of db (as long as I do not write anything is virtual)
$ db 

# See the collections inside that db
$ show collections 

# Returns the data of the collection
$ db.hello.find()

# Delete the collection
$ db.hello.drop()

# Delete db
$ db.dropDatabase()

# Collection = set of json objects
# There are no data constraints ({"age": "18"}, {"age": 18}), there are two
# objects in the same collection with different values for the same key => more flexible.
# a key can have numeric values, strings, bool, but also arrays (ordered list of objects
# without constraint of what these objects are)

# In MongoDb it is listed inside the object, the list of other objects to which it is connected. Joints are almost never made
# because the data is all in the same collection.

# If the key is missing in an object, it is as if it were null

# To create a JSON file directly from the database, use mongoexport
# [go to current dir] 
$ mongoexport --db [db_name] --collection [collection_name] --jsonArray --out [json_name].json

# ------ AGGREGATION FRAMEWORK ------

# Aggregation framework => aggregate: transform data,
# via a pipeline of operations, which must be compatible with each other
# the i.thest must be compatible with the i-1.exima.
# The output could be a: 1.cursor, 2.new collection mongo db