Analysis Module: swepy.analysis

The Analysis class is meant to help get you up and running with quick visualization of various basic SWE metrics, like onset of complete melt by pixel.

class swepy.analysis.Analysis(start_date, swe)

Bases: object

count_melt_onset()

Count the date that each pixel reaches zero for first time of season on a given date. Useful for comparison between years of a given region.

Helper function to manage multi-processing pool for counting melt onset date. Use __count to use a single core

Parralelize on the 2nd axis (spatially)

Returns

df – Count of melt dates of every pixel in image

Return type

pandas DataFrame

create_year_splits()

Take time array from class and create an array of year split indexes

Returns

year_splits – list of julian dates for the start of each year in time series

Return type

list

display_melt_onset_change(dict, year1, year2, interactive=False)

CURRENTLY ONLY WORKS IF FULL YEARS ARE SCRAPED Given a dictionary of melt onset counts by date, viz two years against eachother

Parameters

  • dict (dict) – dictionary of year:list(counts by year) pairs

  • year1 (datetime.date) – first year to plot on bar graph

  • year2 (datetime.date) – second year to plot on bar graph

  • interactive (bool) – Unit testing parameter to help closing plots

Returns

fig – figure for future viz if desired

Return type

matplotlib.figure.Figure

display_summer_change(interactive=False)

Simple built in function for displaying the generated summer change heat map

Parameters

interactive (bool) – boolean switch for unit testing to help close plots (would like to remove)

make_df(time=None, columns=['time', 'count'])

Given a time array, create dateframe with time and count columns

Tool for melt onset workflow

Parameters

  • time (pd.series) – time series of datetime.date() objects

  • columns (list (Optional)) – column names to add in returned dataframe

Returns

df – dataframe with time and count for melt onset analysis

Return type

pd.DataFrame

mask_year_df(df, year)

Mask single year out of pandas dataframe based on “time” column

Parameters

  • df (pd.DataFrame) – dataframe containing a “time” column

  • year (datetime.date) – desired year to get from df

melt_date_year(df)

Grab the counts of each year and stick them in own key of dictionary Makes for easy comparison of melt times each year.

Parameters

df (pd.DataFrame) – dataframe with columns [“time”, “count”]

Returns

counts_dict – dictionary with key=year : value=sum(count in year)

Return type

dict

summer_diff(summer_dict, smooth_cube=None)

Find the average change in summer length by pixel and overall

Parameters

  • summer_dict (dict) – dictionary containing the summer lengths of each year by pixel

  • smooth_cube (np.array(x,x,x)) – swe cube of clean swe data, OPTIONAL

summer_length(smooth_cube)

Function to track summer length for a pixel over each year Generates hash table of key:value = (x,y): summer length in days

Parameters

smooth_cube (np.array) – smoothed (temporally) swe cube

Returns

store – dict using tuple (x,y) as hash key and list of summer lengths by year as key

Return type

dict