Data utilities ============== These items are convenience functions for common data processing work done for pydeck maps. .. autofunction:: pydeck.data_utils.viewport_helpers.compute_view For example, suppose you have 100 points, most of which are centered around London and ten of which are distributed a few 100 kilometers away from it. If you set ``view_proportion=0.9``, pydeck will attempt to fit to the middle 90% of the data, aiming to exclude the points furthest from the core of the visualization. Suppose your ten points are in the Americas in this example--they would be excluded. .. autofunction:: pydeck.data_utils.color_scales.assign_random_colors Examples -------- Below is an illustration of :func:`pydeck.data_utils.color_scales.assign_random_colors`, using a small data set of two classes (here, 'Palace' and 'Clock Tower) and three obervations ('Big Ben', 'Kensington Palace', and 'Buckingham Palace'). Our goal is to quickly color the data by category. >>> import pydeck >>> import pandas >>> data = pandas.DataFrame([ >>> { >>> 'site': 'Big Ben', >>> 'attraction_type': 'Clock Tower', >>> 'lat': 51.5006958, >>> 'lng': -0.1266639 >>> }, >>> { >>> 'site': 'Kensington Palace', >>> 'attraction_type': 'Palace': >>> 'lat': 51.5046188, >>> 'lng': -0.1839472 >>> }, >>> { >>> 'attraction_type': 'Palace', >>> 'site': 'Buckingham Palace', >>> 'lat': 51.501364, >>> 'lng': -0.14189 >>> } >>> ]) >>> color_lookup = pydeck.data_utils.assign_random_colors(data['attraction_type']) >>> # Assign a color based on attraction_type >>> data['color'] = data.apply(lambda row: color_lookup.get(row['attraction_type']), axis=1) >>> # Data now has an RGB color by attraction type: [ { 'site': 'Big Ben', 'attraction_type': 'Clock Tower', 'lat': 51.5006958, 'lng': -0.1266639, 'color': [0, 10, 35] }, { 'site': 'Kensington Palace', 'attraction_type': 'Palace': 'lat': 51.5046188, 'lng': -0.1839472, 'color': [53, 243, 130] }, { 'attraction_type': 'Palace', 'site': 'Buckingham Palace', 'lat': 51.501364, 'lng': -0.14189, 'color': [53, 243, 130] } ]