At Wealthsimple, portfolios have a specified distribution of asset holdings called a target_allocation. This allocation is based on the total value of the asset. For example, a 50/50 allocation between asset A and asset B means the total value of all the asset A you hold is equal to the total value of all the asset B you hold.

When the market price of certain assets change, the account deviates from its target_allocation. When this happens, we can restore the specified balance of assets by buying and selling stocks so that the desired ratio of assets is recovered.

Given a portfolio's target_allocation (ratio of assets), portfolio_holdings (number of each asset), and asset_prices (the cost per unit of the asset), generate a series of trades to ensure the portfolio is balanced according to the target_allocation.

BONUS: We don't want to increase operating costs by doing penny trades. If an asset is within 0.5% of its target allocation, do not trade that asset.

BONUS BONUS: Although Wealthsimple can trade fractions of an asset, not all brokerages can. Apply this restriction to your code and get as close to the target allocation as possible.

You must create a script in Python3, Ruby, or Javascript. This script should read from a file containing test data like the test cases provided, and the name of the file should be configurable.

The output of the script should be a list of trades generated, formatted like the following example:

We placed 3 trades:
1. Bought 120 units of XIC
2. Sold 45 units of ZFM
3. Sold 20 units of VUS

Please fork this repository and make a pull request. It's highly recommended you do this before you start, since you're very likely to run out of time if you attempt to do this at the end.

Do include instructions in your pull request, such as how to configure the test filename and how to run your script.