Computing statistics on sampling periods and target coverage

[oguerrer]

I just uploaded a dataset called big_sample.xlsx. It has more indicators than the ones you have classified so, before processing it, you will need to filter out those indicators that are not part of your classification. Here is a list of the relevant metrics that we want to obtain as part of this data exploration exercise:

• For a given sampling period determine: -- How many indicators do we get? -- How many targets are there? -- How many indicators per target? -- How many SDGs are covered? -- How many many indicators per SDG? -- How many targets per SDG? -- How many missing values would need to be interpolated? -- How many observations do we get before interpolation? -- How many observations would we get after interpolation?

• Obtain these same these same quantities for all sampling periods with at least 7 years of observations.

[daniguariso]

I uploaded the code time_frames_metrics.py that computes the relevant metrics for all possible intervals of seven years. The code also retrieves the interval that maximizes the number of observations before interpolation and the one that maximizes the number of observations after interpolation.