Welcome to the API PT2!

Data Edition

Goal - Filter down to the pit density we want and plot it

Step 1. Imports

# imports
from datetime import date
import geopandas as gpd
import matplotlib.pyplot as plt
from snowexsql.api import PointMeasurements, LayerMeasurements

Step 2. Find the pits in the Boise River Basin

# Find site names we can use
print(LayerMeasurements().all_site_names)

[('Cameron Pass',), ('Sagehen Creek',), ('Fraser Experimental Forest',), ('Mammoth Lakes',), ('Niwot Ridge',), ('Boise River Basin',), ('Little Cottonwood Canyon',), ('East River',), ('American River Basin',), ('Senator Beck',), ('Jemez River',), ('Grand Mesa',)]

# Get the first 1000 measurements from the Boise River Basin Site
df = LayerMeasurements.from_filter(
    type="density",
    site_name="Boise River Basin",
    limit=1000
)

# Explore the pits so we can find an interesting site
df.loc[:, ["site_id", "geom"]].drop_duplicates().explore()

Make this Notebook Trusted to load map: File -> Trust Notebook

Step 3. Pick a point of interest

# We noticed there are a lot of pits (timeseries pits) for Banner Open
# Filter down to ONE timeseries
site_id = "Banner Open"
df = LayerMeasurements.from_filter(
    type="density",
    site_id=site_id
).set_crs("epsg:26911")

df.loc[:, ["site_id", "geom"]].drop_duplicates().explore()

Make this Notebook Trusted to load map: File -> Trust Notebook

# Get the mean of each date sampled
df["value"] = df["value"].astype(float)
df.set_index("date", inplace=True)
mean_values = df.groupby(df.index).mean()
mean_values.head()

	id	latitude	longitude	northing	easting	utm_zone	depth	bottom_depth	value
date
2019-12-18	22609.5	44.30464	-115.23603	4.907222e+06	640699.666121	11.0	26.000000	16.000000	279.000000
2020-01-09	22670.0	44.30463	-115.23601	4.907221e+06	640701.285289	11.0	51.000000	41.000000	156.777778
2020-01-23	22746.0	44.30461	-115.23598	4.907219e+06	640703.725988	11.0	61.727273	51.727273	254.166667
2020-01-30	22830.5	44.30461	-115.23598	4.907219e+06	640703.725988	11.0	72.000000	62.000000	236.000000
2020-02-06	22910.0	44.30458	-115.23594	4.907216e+06	640706.988222	11.0	72.000000	62.000000	254.141026

Notes on this mean

Taking this mean as bulk density could be flawed if layers are overlapping or layers vary in thickness

# !pip install plotly
import plotly.express as px
# For rendering in readthedocs
import plotly.offline as py
py.init_notebook_mode(connected=True)

# Plot the timeseries of mean density
fig = px.line(
    mean_values, x=mean_values.index, y='value',
    title=f'Mean Density - {site_id}',
    labels={'value': 'Density', 'date': 'Date'}
)

fig.update_layout(
    template='plotly_dark'
)

# Show the plot
fig.show()

# alternative matplotlib code
# mean_values["value"].plot()
# plt.title('Mean Density by Date')
# plt.xlabel('Date')
# plt.ylabel('Mean Density')
# plt.gcf().autofmt_xdate()
# plt.show()

# Show more detail by using a box plot
# make sure values are floats
df["value"] = df["value"].astype(float)

# make Make a box plot
fig = px.box(df, x='date', y='value', notched=True, title='Pit Density by Date')
fig.update_traces(hoverinfo='y+name')
fig.update_layout(template='plotly_dark')

# alternative matplotlib code
# df.boxplot(by='date', column='value')
# plt.title('Density Distribution for Banner by Date')
# plt.suptitle('')  # Suppress the automatic title
# plt.xlabel('Date')
# plt.ylabel('Value')
# plt.show()