Python Legend, put legend outside of plot display

Often when you legend is too large, it is convenient to put it off to the side. Here's how...

from matplotlib import pyplot as plt

plt.plot(a,b, label="legend label")
plt.legend(loc='center left', bbox_to_anchor=(1, 0.5),prop={'size':10})

An example here:

Skew-T

Found this python script for plotting skew-t plots. https://pypi.python.org/pypi/SkewT/0.1.1

Python and MesoWest API

Uses the MesoWest API to get ozone concentration data and plot them. An example of formatting plot datetime plot tick marks is also shown.

# Brian Blaylock
# June 9, 2015

# Uses the MesoWest API to get find the maximum ozone
# from in-situ stations during GSLSO3S

# Tips for creating API request:
# -look at MesoWest API documentation: http://mesowest.org/api/docs/
# -use JSON viewer to see request results: http://jsonviewer.stack.hu/


import json
import numpy as np
import urllib2
import matplotlib.pyplot as plt
from datetime import datetime
from matplotlib.dates import DateFormatter, YearLocator, MonthLocator, DayLocator, HourLocator

token = '1234567890' #contact mesowest@lists.utah.edu to get your own token.

station = 'mtmet'
# dateformat YYYYMMDDHHMM
start_time = '201506050000'
end_time = '201506060000'
variables = 'ozone_concentration'
time_option = 'local'
URL = 'http://api.mesowest.net/v2/stations/timeseries?stid='+station+'&start='+start_time+'&end='+end_time+'&vars='+variables+'&obtimezone='+time_option+'&token='+token

#Open URL and read the content
f = urllib2.urlopen(URL)
data = f.read()

# convert that json string into some python readable format
data = json.loads(data)

stn_name = data['STATION'][0]['NAME']
# get ozone data convert to numpy array
ozone = data["STATION"][0]["OBSERVATIONS"]["ozone_concentration_set_1"]
#convert ozone into a numpy array: setting dtype=float replaces None value with a np.nan
ozone = np.array(ozone,dtype=float)

# get date and times and convert to datetime and put into a numpy array
dates = data["STATION"][0]["OBSERVATIONS"]["date_time"]
DATES = np.array([]) # first make an empty array
for i in dates:
   if time_option=='utc':
      converted_time = datetime.strptime(i,'%Y-%m-%dT%H:%M:%SZ')
   else:
      converted_time = datetime.strptime(i,'%Y-%m-%dT%H:%M:%S-0600')
   DATES = np.append(DATES,converted_time)


# make a simple ozone time series
#----------------------------------------------------------
ax = plt.subplot(1,1,1)
plt.plot(DATES,ozone)
plt.title('Ozone Concentration (ppb) for '+stn_name)
plt.xlabel('date')
plt.xticks(rotation=30)
#Now we format the date ticks
# Format Ticks
# Find months
months = MonthLocator()
# Find days
days = DayLocator()

# Find each 0 and 12 hours
hours = HourLocator(byhour=[0,6,12,18])
# Find all hours
hours_each = HourLocator()
# Tick label format style
dateFmt = DateFormatter('%b %d\n%H:%M')
# Set the x-axis major tick marks
ax.xaxis.set_major_locator(hours)
# Set the x-axis labels
ax.xaxis.set_major_formatter(dateFmt)
# For additional, unlabeled ticks, set x-axis minor axis
ax.xaxis.set_minor_locator(hours_each)


plt.savefig(start_time+'_'+end_time+'.png', format='png')

Putty Change color scheme

I've been staring at the hard-to-see blue color too long in Putty. I found this helpful info for changing the default color scheme in python here...http://www.darkrune.org/blog/?p=213. We'll see how I like the change.

Before logging into a Putty session, click Window/Colours, and reassign the colors

Putty RGB colors/options for the Zenburn color scheme are as follows -

• Default Foreground - 255/255/255
• Default Background - 51/51/51
• ANSI Black - 77/77/77
• ANSI Green - 152/251/152
• ANSI Yellow - 240/230/140
• ANSI Blue - 205/133/63
• ANSI Blue Bold -135/206/235
• ANSI Magenta - 255/222/173 or 205/92/92
• ANSI Cyan - 255/160/160
• ANSI Cyan Bold - 255/215/0
• ANSI White - 245/222/179

Bootstrapping in Python

Python code for regression significance testing

#------------------------------------------
#Bootstrap test for significance
#------------------------------------------
import numpy as np
import matplotlib.pyplot as plt

x = compositeL # some numpy vector
y = compositeF # another numpy vector

# we will fill this vector with the regression coefficient as we find them
b1_vector = np.zeros(5000)

#bootstraping
for i in np.arange(5000):
     index = np.random.randint(0,len(x),len(x))
     #pull out the (x,y) pair for each index
     sample_x = x[index]
     sample_y = y[index]
     #calculate the b1 value and store it in the b_vector
     b_values = np.polyfit(sample_x,sample_y,1)
     #b1 is the first element result of polyfit 
     b1=b_values[0]
     b1_vector[i]=b1

plt.figure(6)

plt.hist(b1_vector,50)

In this example there is 99.96% confidence that the regression coefficient is positive.