1. Compute correlations between time series
This case study focused a lot on visualizing and understanding the statistical properties of the time series data in the jobs dataset. Throughout these exercises, you noticed and discussed a number of interesting patterns in our data. For example, you saw how the financial crisis of 2008 led to a rise in unemployment rates across all industries. You were also able to reveal that certain industries were more affected by seasonality, as we saw that the Agriculture and Construction industries saw rises in unemployment rates during the colder months of winter. In the following set of exercises, you will apply the principles that you have learned about correlations in time series in order to uncover which job industries are the most similar or the most different.
2. Trends in Jobs data
We will be making use of the trend_df DataFrame from the last video.
3. Plotting a clustermap of the jobs correlation matrix
First, compute the correlation between all columns in the trend_df DataFrame using the spearman method and assign the results to a new variable called trend_corr. Then, generate a new clustermap() of the correlation matrix trend_corr by using the clustermap function from the seaborn library. Lines 3 and 4 specify a rotation angle of 0 to the y-axis labels, and a rotation angle of 90 to the x-axis labels.
4. The jobs correlation matrix
Now let's try some examples.
5. Let's practice!