In [1]:
import folium
import pandas as pd
url = (
"https://raw.githubusercontent.com/python-visualization/folium/master/examples/data"
)
state_geo = f"{url}/us-states.json"
state_unemployment = f"{url}/US_Unemployment_Oct2012.csv"
state_data = pd.read_csv(state_unemployment)
m = folium.Map(location=[48, -102], zoom_start=3, tiles="Stamen Toner")
folium.Choropleth(
state_geo,
data=state_data,
columns=["State", "Unemployment"],
key_on="feature.id",
fill_color="YlGn",
fill_opacity=0.7,
line_opacity=0.2,
legend_name="Unemployment Rate (%)",
).add_to(m)
popup = "Must be on top of the choropleth"
folium.CircleMarker(
location=[48, -102],
radius=10,
fill=True,
popup=popup,
weight=1,
).add_to(m)
m
Out[1]:
Basic datatypes¶
In [1]:
x=1
y=2
In [2]:
print(x)
In [5]:
x = 16 # integer
y = 2.6 # float
z = 'abc' # string
# this is a comment
In [4]:
print(z)
In [7]:
result = x+y
In [8]:
print(result)
In [9]:
365*10
Out[9]:
In [10]:
365*24
Out[10]:
In [11]:
a = True
In [12]:
not a
Out[12]:
Other useful python types¶
In [13]:
# list
l = [1, 2, 3, 'abc', 3.5, True, [4,5,6]]
In [14]:
# indexing starts at 0!
l[3]
Out[14]:
In [15]:
# tuple
t = (1,2,3)
In [16]:
t[1]
Out[16]:
In [17]:
# tuples are like lists, but lists are more flexible
l.append(89.7)
In [18]:
l
Out[18]:
In [19]:
# the syntax object.method() is very common
l.reverse()
In [20]:
l
Out[20]:
In [21]:
# dictionary
d = {'a': 1, 'b': 2}
In [22]:
d['a']
Out[22]:
In [23]:
colors = ['red', 'green', 'blue', 'black']
d['red'] = (255, 0, 0)
d['blue'] = (0,0,255)
In [24]:
d['red']
Out[24]:
In [25]:
d
Out[25]:
In [26]:
list(d.keys())
Out[26]:
In [27]:
list(d.values())
Out[27]:
Numpy¶
In [28]:
import numpy as np
In [31]:
a = np.array([0,1,2,3,4,5])
In [38]:
# an example of method access
a.max()
Out[38]:
In [39]:
# ndim is one of the attributes of a numpy array.
a.ndim
Out[39]:
In [40]:
a.shape
Out[40]:
In [41]:
b = np.array([[0,1,2,3,4], [5,6,7,8,9]])
In [42]:
b.ndim
Out[42]:
In [44]:
b
Out[44]:
In [45]:
b.shape
Out[45]:
In [46]:
# The vertical axis is always the first
# This is second row, third column
b[1,2]
Out[46]:
In [47]:
# This is a mistake
b = np.array([[0,1,2,3,4], [5,6,7,8]])
In [50]:
my_zeros = np.zeros((2000,2000))
In [53]:
# top corner
my_zeros[0,0] = 1
my_zeros
Out[53]:
In [54]:
# set first row to 1
my_zeros[0,:] = 1
my_zeros
Out[54]:
In [55]:
# set 3 first elements of second row to 2
my_zeros[1,0:3] = 2
# equivalent is my_zeros[1,:3]
my_zeros
Out[55]:
In [56]:
# set last 2 elements of first column to 6
my_zeros[-2:, 0] = 6
my_zeros
Out[56]:
In [58]:
# set a square 100x100 with top corner at (350, 450) to 10.
my_zeros[350:450, 450:550] = 10
In [60]:
my_piece_of_my_zeros = my_zeros[:2,:2]
In [61]:
my_piece_of_my_zeros
# Tricky: this is a "view" on the original array, not a copy
Out[61]:
In [62]:
my_piece_of_my_zeros[0,0] = 5
my_piece_of_my_zeros
Out[62]:
In [63]:
my_zeros
Out[63]:
In [82]:
my_zeros.dtype # float64 - also known as double-precision float (8 byte = 64 bit)
Out[82]:
In [64]:
# Typical mistake:
first = np.zeros((100,100))
second = first[:40,:50]
first[0,0] = 1
# This would modify 'second' as well!
In [65]:
# solution
first = np.zeros((100,100))
second = first[:40, :50].copy()
# OR
second = first[:40, :50] + 0.
In [66]:
# operations on arrays
first = np.zeros((100,100))
second = np.ones((100,100))
In [69]:
first*second
Out[69]:
In [70]:
i, j = np.indices((100,100))
In [71]:
i
Out[71]:
In [72]:
j
Out[72]:
In [73]:
i[34, 78]
Out[73]:
In [74]:
j[34,78]
Out[74]:
In [75]:
(i - 50)
# (in place operation )
# i -= 50
# result is same as i = i - 50
Out[75]:
In [76]:
j - 50
Out[76]:
In [77]:
r = np.sqrt((i-50)**2 + (j-50)**2)
In [81]:
r[48:52,48:52]
Out[81]:
In [ ]:
# Some more discussion about memory
r0 = r
r += 1 # will apply to r0
r = r+1 # Now r is a new object different from r0
