I am relatively new to pandas and have a dataset which requires categorising the values by a subset (the year) of the dataset in pandas.

The dataset is one where there are individual rows for each reported date and I need to bin and categorise each of them into eight categories.

The current project is here: Scottish Crime Pandas Project where you can find the current notebook and parquet of where the dataset is currently.

Here is a sample of the dataset

The date_start column is to be used to bin and categorise the data and we are looking to bin based on the count_per_10k column.

I have successfully categorised each row using np.linspace and pd.cut for the whole historic dataset. This has been done by:

count_10k_bins = np.linspace(min(df_crime['count_per_10k']), max(df_crime['count_per_10k']), 8)

count_10k_bins_names = ['Minimal', 'Very Low', 'Low', 'Medium', 'High', 'Very High', 'Extremely High']

df_crime['count_category_historic'] = pd.cut(df_crime['count_per_10k'], 
                                             count_10k_bins, 
                                             labels=count_10k_bins_names, 
                                             include_lowest=True)

Please see this image for a sample of the dataset – Datset Example

I cannot work out a way to categorise each individual row in the dataset based on the subset of the year it falls in