Datecode not correct for "stocks".csv . (unsupported operand for "-")

[minertom]

This is probably an easy fix to correct.

In the "cvxpy-stocks.ipynb" notebook file, which I have downloaded locally, I am using the stocks file "bigmac.csv", to import my stock porfolio.

The line:
df_returns = df_stocks.set_index('date').diff()

Returns the error
TypeError: unsupported operand type(s) for -: 'str' and 'str'

I did try different permutations of the code, and interestingly, one version worked and another one did not.

The code in the example is close to this:
df_stocks = pd.read_csv("bigmac.csv").iloc[500:1800]
df_returns = df_stocks.set_index('date').diff()
mean_stock = df_returns.mean().values

It results in the error
TypeError: unsupported operand type(s) for -: 'str' and 'str'

However, the following code works:
df_stocks = pd.read_csv("bigmac.csv").iloc[500:1800]
df_returns = df_stocks.set_index('date')
differences = df_returns.diff' #**Note that diff here has no () following**
mean_stock = df_returns.mean()
values = mean_stock.values
print(values)

And the results are
[2029.88108208 723.91574523 3.5758786 ]

So, what is the difference between the two versions and why does yours work?

Thank You
Tom Cipollone