CODEWITHSUNDEEP
pythonpandasdataframe
H.Choi
H.Choi

Reputation: 3223

Generating future dataframe with annual growth rate

I am trying to generate annual data for a certain product when I have data for the base year and growth rate.

In the toy example, each product has different annual growth rate in efficiency by its 'color', and I want to generate yearly data until 2030.

Therefore, I have base year data (base_year) as follows:

    year    color   shape    efficiency
0   2018    red     circle   50
1   2018    red     square   30
2   2018    blue    circle   100
3   2018    blue    square   60

And each type of product's growth rate (growthrate) as: 

    color   rate
0   red     30
1   blue    20

Results I desire is:

    year    color   shape    efficiency
0   2018    red     circle   50
1   2018    red     square   30
2   2018    blue    circle   100
3   2018    blue    square   60
4   2019    red     circle   65
5   2019    red     square   39
6   2019    blue    circle   120
7   2019    blue    square   72
8   2020    red     circle   84.5
... (until 2030)

The data used in the toy code is..

base_year = pd.DataFrame(data = {'year': [2018,2018,2018,2018],
                     'color': ['red', 'red', 'blue', 'blue'],
                     'shape' : ['circle', 'square', 'circle', 'square'],
                     'efficiency' : [50, 30, 100, 60]}, columns = ['year', 'color', 'shape', 'efficiency'])

growthrate = pd.DataFrame(data = {'color': ['red', 'blue'],
                     'rate' : [30, 20]}, columns = ['color', 'rate'])

I've been trying some approach using .loc, but it seems such approach is quite inefficient.

Any suggestions or hints would be appreciated. Thank you in advance!

Upvotes: 1

Views: 98

Answers (1)

Chris Adams
Chris Adams

Reputation: 18647

Here is one way to do this:

years = 2031 - 2018

df = (pd.concat([df.assign(year=df['year']+i,
                          efficiency=df['efficiency']*((df['rate']/100+1)**i)) 
                for i, df in enumerate([base_year.merge(growthrate, on='color')] * years)])
      .drop('rate', axis=1))

Upvotes: 2

Related Questions