Skip to content

Latest commit

 

History

History
132 lines (111 loc) · 4.33 KB

notes.org

File metadata and controls

132 lines (111 loc) · 4.33 KB
Raw

introduction [30 minutes]

slides

UI tools

  • if you have used the pan / zoom tools or the ‘g’, ‘l’, ‘k’, etc keys in mpl, you have used the mpl event framework
    • show demo of keys / pan / zoom
    • the updating coordinates
  • events has always been in mpl (Event class came into codebase on Tue Dec 28 2004 as a refactoring of existing functionality

There is more than a book of material here

Interactive applications using matplotlib by Ben Root

“Learn how Lincoln could use Matplotlib to save the Earth from hordes of ravenous space ware-rabbits! Just plotting your data won’t save the world, but turning your plots into interactive applications will help your users explore their data and gain greater insight into their problems. This book will walk the reader through the process of turning a simple radar viewer into a useful hazardous event tracker, so you too can spot where the space ware-rabbits are going to strike next!” - Ben Root

https://www.packtpub.com/application-development/interactive-applications-using-matplotlib

install

$ conda create mpl_tut -c anaconda matplotlib pandas pytables ipython h5py scipy python=3.6 $ source activate mpl_tut

00-explore [5 + 15 minutes]

  • ‘hello world’ of event processing
  • layers of mpl
    • figure / axes / artist vs canvas distinction
    • figure vs figure manager vs canvas
  • canvas.mpl_connect
  • canvas.mpl_disconnect
  • Three ‘classes’ of events
    • active
      • [‘button_press_event’, ‘button_release_event’, ‘scroll_event’,

    ‘key_press_event’, ‘key_release_event’, ‘pick_event’]

    • passive
      • [‘motion_notify_event’, ‘figure_enter_event’, ‘axes_enter_event’, ‘axes_leave_event’]
    • internal
      • [‘draw_event’, ‘resize_event’]
      • important if you are doing blitting (which is it’s own tutorial)
  • use OO interface when ever possible

01-callable [

  • frequently want to carry state between user interactions
  • simplest way (short of globals) is a callable class
  • going to want to restart IPython after this
    • or `matplolib.rcdefaults()`

WARNING

  • the callback registry (matplotilb.cbook.CallbackRegistry) only hold weakrefs to the callback functions. If you do not explicitly keep a ref to them around they maybe garbage collected.
  • this is probably the biggest gotcha
  • done to prevent reference cycles

02-event filtering

  • use an object with several methods, hook them all up to events
  • look at the values in the events to decide what to do
  • hold an artist as state and mutate on user input

interlude about temperature data

  • data sets from NOAA, historical hourly weather reports
  • only extracting the air temperature, a LOT more information in the reports
  • raw data extraction code in 99-collect_data.py, but processed data from BWI in repo
  • ~10 years of ~hourly air temperature readings
from w_helpers import load_data

bwi = load_data('nyc')
  • plot full time series
    • this is sort of useless, but can zoom in!
  • times are in UTC because I lost to timezone normalization while prepping for this

03-picking

  • look at 2015-01-01 to a few days ago aggregated by day
    • look at bwi_daily columns
  • looking a bit better, but only looking at the mean
    • can see some interesting days
  • ‘pick_event’ can have arbitrary attributes added by the picker
    • show `set_picker` docstring
  • `set_gid` / `get_gid` a place for users to stash an id, not used my mpl
  • `set_label`/`get_label` are the name used for automatic legend generation

04-custom_plotting

  • mpl’s ‘native’ API is low level
  • write the API you need for your data (inside of which you bring your data to the API)
  • i-haz-a-soap-box

05-interactive_temperature

  • fully functioning ‘app’
  • what else could you make it do?

99-get data

  • $ conda install -c conda-forge cartopy proj4
  • click on the stations to print station name + template
  • use

Examples to look at

event_handling

  • data_browser.py
  • legend_picking.py
  • path_editor.py
  • poly_editor.py
  • viewlims.py
  • zoom_window.py

widgets

  • buttons.py
  • check_buttons.py
  • cursor.py
  • lasso_selector_demo_sgskip.py
  • polygon_selector_demo.py
  • rectangle_selector.py
  • slider_demo.py
  • span_selector.py