iridl / python-maproom-tuto

Python Maprooms that are also a Tutorial to learn how to make them

python-maproom-tuto

Python Maprooms that are also a Tutorial to learn how to make them.

We are building Maproom Applications with Dash and with Dash-Leaflet for the mapping components.

We are performing data analysis with Xarray, NumPy and pandas

The Tutorial reviews the following Lessons. To get started, after having installed conda, create the tutorial environment by running:

conda create -n mr-tuto --file conda-osx-64.lock

(substituting osx or win for linux as appropriate)

and then activate it. Discover Lesson 1 by running the application with:

CONFIG=config-sample.yaml python maproom.py

• Lesson 1 : Titles and text

• Lesson 2 : Topo background map

• Lesson 3 : Add Street background map

• Lesson 4 : Set a Variable Dropdown menu

• Lesson 5 : Add a data layer to the map

• Lesson 6 : Automatically center map on data

• Lesson 7 : Prevent dragging far from data

• Lesson 8: Add the colorscale

• Lesson 9: Tick colorscale every 10

• Lesson 10 : Configure data source

• Lesson 11 : Use other data

• Lesson 12: Add another variable: temperature

• Lesson 13 : Adjust temperature colorscale

• Lesson 14 : Map Climatology for a given month

• Lesson 15 : Callback for Map Title

• Lesson 16 : Location Selection

• Lesson 17 : Plot Local Climatology

python-maproom-tuto's Issues

Lesson 5 need be broken up in many sub-lessons

e.g.:

• Read data
• rename dimensions
• select one point in time
• add callback in maproom (can we make the tile without a callback?)
• add data in callback
• add prefix in config and maproom
• add overlay in layout

I lost almost all participants on Lesson 5 :(

New Lesson: a line plot for temperature

as opposed to bars that can be kept for precipitation

Lesson 3 needs more details

In particular the tiling suffixes were too much of a mystery.

Have Lesson 6 starts with hard-coding it

then progressively have it automated. It's better because then one see the end results of the purpose of the lesson at its beginning.

New Lessons: add errors handling

For users as well as for developers

New Lesson: mouse hover map reveals admin's name

for the finest admin resolution

Have Lesson 15 before Lesson 5

In order to introduce callbacks in a simple manner. Here is Lesson 5 it was a killer. Unless the tiled data map can be added without callback.

A Lesson on having the variables in config

Listing their names, long_name, description, file path, etc...

Caluclations need be explicit lessons as well

Can't just say: compute climatology because likely won't know Xarray, and not even Python. Also need to give instructions on how to creating a dummy .py file to run code and print stuff to check on calculations. Would be nice to point them to the documentation of functions you'd want them to use to realize a given calculation, even if in the Zambia case, it would not have helped since they had no internet access, thus no access to documentation. But maybe we can assume this is a very specific case...

New Lesson: A title for local plot

including printing of location

New Lesson: overlay admin layers