Python 3 - indexing Dataframe using a float type variable

Question

I need to subset a particular column from a Dataframe, of simulated stock prices, and find its mean.

Variables previously defined are:

T  = 1  
dt = 1/1000 
which makes T/dt = 1000. (float)

Now, directly indexing DataFrame as follows, throws an error:

StockPrice[T/dt].mean() -> error

However, casting index as 'int' before using, works fine:

StockPrice[int(T/dt)].mean()

So I am trying to understand, what is the standard practice when sub-setting DataFrames using other variables that may generate integer values (but with float datatype). Should we cast them as int and then use them, or is there an alternate way?

Which version of pandas are you using? What is the exact error message? — albert
– albert, Commented Feb 15, 2016 at 18:43
Pandas Version: 0.17.1 Snippet from error message: 1974 # get column 1975 if self.columns.is_unique: 1976 return self._get_item_cache(key) 1977 1978 # duplicate columns & possible reduce dimensionality KeyError: 1000.0 — Uditg_ucla
– Uditg_ucla, Commented Feb 15, 2016 at 18:51

Dickster · Accepted Answer · 2016-02-16 11:37:10Z

2

Given the stock prices are a continuous variable you might be best to use range to capture relevant stock prices around your target price. That range can be as large or as small as needed.

Pandas Series comes with a .between() method. This will evaluate to True or False for each value of the series within the range. Then use this 'criteria' in a boolean slicing operation to pull out the relevant values.

np.random.seed(1)

df = pd.DataFrame(np.random.rand(1000,1),columns=['stockprice'])*10000.

epsilon = 100.
dt = 1000.

criteria = df['stockprice'].between(dt-epsilon,dt+epsilon)
print df[criteria]

edited Feb 16, 2016 at 11:37

answered Feb 15, 2016 at 21:56

Dickster

3,0193 gold badges25 silver badges29 bronze badges

Sign up to request clarification or add additional context in comments.

1 Comment

Ajean Over a year ago

While this code may answer the question, providing additional context regarding why and/or how this code answers the question improves its long-term value.

Stop harming Monica · Accepted Answer · 2016-02-15 22:39:21Z

0

You should. I get this warning in pandas 0.17.1:

FutureWarning: scalar indexers for index type Int64Index should be integers and not floating point

It's a feature, not a bug.

Besides, using floats seems to work for Series but not with DataFrames so the future is almost here.

answered Feb 15, 2016 at 22:39

Stop harming Monica

12.7k1 gold badge40 silver badges63 bronze badges

1 Comment

Uditg_ucla Over a year ago

Thanks this is helpful!

Collectives™ on Stack Overflow

Python 3 - indexing Dataframe using a float type variable

2 Answers 2

1 Comment

1 Comment

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

1 Comment

1 Comment

Your Answer

Sign up or log in

Post as a guest

Related