Single station analysis

Import the familiar modules and the ACIS data loader



In [1]:

    
import ACISLoader as A
import pandas as pd
import matplotlib.pyplot as plt

Define your request parameters:

COOP station ID 304174 is Ithaca, NY
por indicates the period of record
get minimum temperature, maximum temperature and daily precipitation

The loader returns a panel with a dataframe keyed on the station id.



In [2]:

    
p = A.ACISLoader(sid='304174',sdate='2012-01-01',edate='por',elems='mint,maxt,pcpn')
df = p['304174']

The dataframe has a column per requested element, indexed by the date



In [3]:

    
df[-10:]

Use pandas and matplotlib to visualize the min/max temperatures



In [4]:

    
a1 = df.mint.plot(figsize=(16,6)).set_title('Ithaca, NY min and max temperature')
a2 = df.maxt.plot(style='r')

Use pandas to smooth the data



In [5]:

    
b1 = pd.rolling_mean(df.mint,10).plot(figsize=(16,6)).set_title('10 day moving average')
b2 = pd.rolling_mean(df.maxt,10).plot(style='r')

Use ACIS to calculate the moving average

10 days ending on reported date



In [6]:

    
p1 = A.ACISLoader(sid='304174',sdate='2012-01-01',edate='por',
                 elems=[dict(name='mint',interval='dly',duration=10,reduce='mean'),
                        dict(name='maxt',interval='dly',duration=10,reduce='mean')])
df1 = p1['304174']
a1 = df1.mint.plot(figsize=(16,6)).set_title('Ithaca, NY min and max temperature (10-day mean)')
a2 = df1.maxt.plot(style='r')

Don't use pandas to sum the precipitation

The ACIS `sum` reduction will handle subsequent/accumulated values properly

Get the 120 years of precipitation over 30, 60, 90 and 120 days ending on September 1.



In [7]:

    
p = A.ACISLoader(sid='304174',sdate='1893-09-01',edate='por',elems=[
    dict(label='p_30',name='pcpn',interval=(1,0,0),duration=30,reduce='sum'),
    dict(label='p_60',name='pcpn',interval=(1,0,0),duration=60,reduce='sum'),
    dict(label='p_90',name='pcpn',interval=(1,0,0),duration=90,reduce='sum'),
    dict(label='p_120',name='pcpn',interval=(1,0,0),duration=120,reduce='sum'),
    ])
prcp = p['304174']



In [8]:

    
fig, axes = plt.subplots(nrows=2, ncols=2, figsize=(16, 6),sharex=True,sharey=True)
prcp.p_30.hist(ax=axes[0,0],color='k', alpha=0.5, bins=50); axes[0,0].set_title('30-day')
prcp.p_60.hist(ax=axes[0,1],color='k', alpha=0.5, bins=50); axes[0,1].set_title('60-day')
prcp.p_90.hist(ax=axes[1,0],color='k', alpha=0.5, bins=50); axes[1,0].set_title('90-day')
prcp.p_120.hist(ax=axes[1,1],color='k', alpha=0.5, bins=50); axes[1,1].set_title('120-day')









    Out[8]:





<matplotlib.text.Text at 0x106261e10>



In [8]:

	maxt	mint	pcpn
2013-06-18	80	54	0.14
2013-06-19	67	41	0
2013-06-20	67	42	0
2013-06-21	79	51	0
2013-06-22	83	61	0
2013-06-23	86	66	0
2013-06-24	90	66	0.81
2013-06-25	85	63	0.06
2013-06-26	83	65	0.03
2013-06-27	82	60	0

Single station analysis

Import the familiar modules and the ACIS data loader

Define your request parameters:

The loader returns a panel with a dataframe keyed on the station id.

The dataframe has a column per requested element, indexed by the date

Use pandas and matplotlib to visualize the min/max temperatures

Use pandas to smooth the data

Use ACIS to calculate the moving average

Don't use pandas to sum the precipitation

The ACIS sum reduction will handle subsequent/accumulated values properly

The ACIS `sum` reduction will handle subsequent/accumulated values properly