Ironhack Logo

MongoDB | Compass CRUD

Introduction

We are back with our queries! 😉

We have learned some super useful query operators, that will helps us to make much better queries to retrieve the data we need. We will continue using the Crunchbase database we used on the last exercise.

Deliverables

Since we will be querying our database from Mongo Compass, you will need to copy/paste the query, projection, sort, skip and limit you entered on Mongo Compass. On the queries.md file, you will find the instructions about the queries you need to do, and a field to fill the answers.

Example

  1. This is an example
  • query: /You should copy/paste the query in here/
  • projection: /You should copy/paste the projection in here/
  • sort: /You should copy/paste the sort in here/
  • skip: /You should copy/paste the skip in here/
  • limit: /You should copy/paste the limit in here/

Instructions

Iteration 1

We already imported the database for the last exercise, but just in case you didn't, follow the instructions. In case you already have the database, jump to the second iteration.

First, we need to import the database we will be using for the lab. We will use the Crunchbase database. Crunchbase is the premier destination for discovering industry trends, investments, and news about hundreds of thousands of companies globally. From startups to Fortune 500s, Crunchbase is recognized as the primary source of company intelligence by millions of users globally.

The database contains more than 18k documents, and each of them has a lot of information about each of the companies. A document looks like the following:

image

  1. You will find the .zip file of the Database on the lab folder.
  2. Unzip the file
  3. From the terminal, import the database to Mongo using the following command:
$ mongoimport --db companies --collection companies --file companies.json
  1. Check on Mongo Compass if everything goes ok:

image

:::info When running the mongoimport you should be located in the same folder the companies.json file. :::

Iteration 2

You already know how this goes, so let's start working:

  1. All the companies that it's name match 'Babelgum'. Retrieve only their name field.
  2. All the companies that have more than 5000 employees. Limit the search to 20 companies and sort them by number of employees.
  3. All the companies founded between 2000 and 2005, both years included. Retrieve only the name and founded_year fileds.
  4. All the companies that had an IPO of more than 100.000.000 and have been founded before 2010. Retrieve only the name and ipo fields.
  5. All the companies that have less than 1000 employees and have been founded before 2005. Order them by the number of employees and limit the search to 10 companies.
  6. All the companies that don't include the partners field.
  7. All the companies that have a null type of value on the category_code field.
  8. All the companies that have at least 100 employees but less than 1000. Retrieve only the name and number of employees fields.
  9. Order all the companies by their IPO price descendently.
  10. Retrieve the 10 companies with more employees, order by the number of employees
  11. All the companies founded on the second semester of the year. Limit your search to 1000 companies.
  12. All the companies that have been 'deadpooled' after the third year.
  13. All the companies founded before 2000 that have and acquisition amount of more than 10.000.000
  14. All the companies that have been acquired after 2015, order by the acquisition amount, and retrieve only their name and acquisiton field.
  15. Order the companies by their founded year, retrieving only their name and founded year.
  16. All the companies that have been founded on the first seven days of the month, including the seventh. Sort them by their aquisition price descendently. Limit the search to 10 documents.
  17. All the companies on the 'web' category that have more than 4000 employees. Sort them by the amount of employees in ascendant order.
  18. All the companies which their acquisition amount is more than 10.000.000, and currency are 'EUR'.
  19. All the companies that have been acquired on the first trimester of the year. Limit the search to 10 companies, and retrieve only their name and acquisition fields.
  20. All the companies that have been founded between 2000 and 2010, but have not been acquired before 2011.

Good luck! 😉