In [ ]:
...

root¶

Progressive Loading and Visualization¶

This ProgressiBook demonstrates the use of chaining widgets to implement the scenario introduced in userguide1.1.ipynb. It progressively downloads New York Yellow Taxi trip data and visualizes the pickup locations.

In [1]:
...

Taxis¶

This first widget handles the configuration and progressive loading of CSV files. It utilizes the sniffer available in the CSV loader for more details.

In [2]:
Constructor.widget('Taxis', 0)
Out[2]:

Quantiles¶

This widget creates a Quantiles module and connects it to the output of the previously defined CSV loader. The Quantiles module maintains an internal data structure, known as a data sketch, to quickly (though approximately) compute quantiles for all loaded numerical columns. This approach is necessary because the dataset's minimum and maximum values are noisy. At this stage, the Quantiles module is created and partially configured: both the pickup_latitude and pickup_longitude columns are selected (although the incomplete snapshot below only displays pickup_longitude).

In [3]:
Constructor.widget('Quantiles', 0)
Out[3]:

Heatmap¶

This widget finalizes the configuration of the min and max values for the previously created Quantiles. It also creates the Histogram2D module to count all pickup locations on a 512x512 grid. Finally, a Heatmap module is created and connected to the output of the Histogram2D module. This module converts the 2D histogram into an image, displayed below.

In [4]:
Constructor.widget('Heatmap', 0)
Out[4]: