�ݺ�ߣ

Geospatial
Visualization
GeoPandas

2
What is Geopandas?
 GeoPandas is an open-source project to make
working with geospatial data in python easier.
 GeoPandas is one of the most satisfying Python
packages to use because it produces a tangible,
visible output that is directly linked to the real world.
 Here we will be exploring the method to create a geo
map and visualize data over it, using shapefiles(.shp)
and some other Python libraries.

3
The core data structure in GeoPandas
 GeoDataFrame, a subclass of pandas.
 DataFrame, that can store geometry columns and perform
spatial operations.
 The geopandas.GeoSeries, a subclass of pandas.Series,
handles the geometries.
 Therefore, your GeoDataFrame is a combination of
pandas.Series, with traditional data (numerical, boolean, text
etc.), and geopandas.GeoSeries, with geometries (points,
polygons etc.).
 You can have as many columns with geometries as you wish;
there’s no limit typical for desktop GIS software.

4

5
 Each GeoSeries can contain any geometry type (you can even
mix them within a single array) and has a GeoSeries.crs
attribute, which stores information about the projection (CRS
stands for Coordinate Reference System).
 Therefore, each GeoSeries in a GeoDataFrame can be in a
different projection, allowing you to have, for example,
multiple versions (different projections) of the same geometry.
 Only one GeoSeries in a GeoDataFrame is considered the
active geometry, which means that all geometric operations
applied to a GeoDataFrame operate on this active column.

6
world map projections.
 1. Mercator
 This projection was developed by Gerardus Mercator back in 1569
for navigational purposes.
 Its ability to represent lines of constant course from coast to coast
made it the perfect map for sailing the seas.
 Its popularity was so great that it became used as a geographic
teaching aid even though the projection grossly distorts countries
sizes.
 This is at its worst the closer you are to the poles. Greenland is 550%
too big, it should fit into Africa 14 times!

7
Vector Data: OpenGIS Simple Features / OCG Simple Feature Access
Used in spatial databases (Postgresql/PostGIS), GIS, …
WKT (Well known Text):
• Point, Multipoint
• LineString, MultiLineString
• Polygon, MultiPolygon
• GeometryCollection
• (TIN, Circle, Curve, Triangle, Surface, PolyhedralSurface,
…)
map projections.

10
world map projections.
 1. Mercator
 This projection was developed by
Gerardus Mercator back in 1569 for
navigational purposes.
 Its ability to represent lines of constant
course from coast to coast made it the
perfect map for sailing the seas.
 Its popularity was so great that it
became used as a geographic teaching
aid even though the projection grossly
distorts countries sizes.
 This is at its worst the closer you are to
the poles. Greenland is 550% too big, it
should fit into Africa 14 times!

11
Map projections.
 2. Robinson
 This map is known as a ‘compromise’, it shows neither the shape or land
mass of countries correct.
 Arthur Robinson developed it in 1963 using a more visual trial and error
development.
 Our Classic world map uses the Robinson projection and is a contemporary
tribute to the familiar schoolroom map and is perfect for map-lovers of all
ages..

12
map projections
 3. Dymaxion Map
 This projection was released by R Buckminster Fuller in 1954 after several
decades working on it.
 The world is projected onto the surface of a icosahedron, allowing it to be
unfolded and flattened in two dimensions.
 It is said to represent the Earth’s continents as “one island”.

13
map projections
 4. Gall-Peters
 This is a cylindrical world map projection, that regains accuracy in surface
area.
 It is named after James Gall and Arno Peters. Whilst Gall, first described
the projection in 1855 it was not until 1973 when Peters, began to heavily
market the projection as the ‘Peters World Map’ that it became popular.

14
map projections
5. Sinu-Mollweide
 Developed in 1953 by Allen K Philbrick, this projection fuses the
Sinusoidal projection , which was first used in the 16th Century, with Karl
Brandan Mollweide's map of 1805 and challenges our assumption of how
the flattened globe should look.
 Still an equal area projection that maintains fidelity of area, we like this
projection for its bold graphic view.

15
map projections
 6 Winkel Tripel
 A globe that is projected onto a flat surface giving it curved lines of latitude and
curved meridians.
 The projection, by Oswald Winkel in1921 was developed with the goal of minimizing
the three kinds of distortion: area, direction and distance.
 Thus it became the Tripel Projection (German for triple).
 The projection is neither equal-area nor conformal, its main feature is that all of the
parallels are curved except for the straight poles and equator.
 This gives a lovely spherical feeling to this two dimensional map.

16
Download Free Shapefile Maps
 https://www.igismap.com/download-free-shap
efile-maps/
 https://www.statsilk.com/maps/download-free-
shapefile-maps
 https://mapscaping.com/download-shapefiles-f
or-any-country/

17
GeoPandas packages:
 pandas
 shapely
 fiona
 pyproj
 packaging
 Further, matplotlib is an optional dependency,
required for plotting.
 GeoPandas using the conda package manager.

18
Reading and writing files
 Assuming you have a file containing both data and geometry
(e.g. GeoPackage, GeoJSON, Shapefile), you can read it using
geopandas.read_file(),
 which automatically detects the filetype and creates a
GeoDataFrame.
 This tutorial uses the "nybb" dataset, a map of New York
boroughs, which is available through the geodatasets package.
 Therefore, we use geodatasets.get_path() to download the
dataset and retrieve the path to the local copy.

19
Reading and writing files
 Assuming you have a file containing both data and geometry
(e.g. GeoPackage, GeoJSON, Shapefile), you can read it using
geopandas.read_file(),
 which automatically detects the filetype and creates a
GeoDataFrame.
 This tutorial uses the "nybb" dataset, a map of New York
boroughs, which is available through the geodatasets package.
 Therefore, we use geodatasets.get_path() to download the
dataset and retrieve the path to the local copy.

20
Writing files
 To write a GeoDataFrame back to file use
GeoDataFrame.to_file().
 The default file format is Shapefile, but you can
specify your own with the driver keyword.

21
Simple accessors and methods
 Now we have our GeoDataFrame and can start working with
its geometry.
 Since there was only one geometry column in the New York
Boroughs dataset, this column automatically becomes the
active geometry and spatial methods used on the
GeoDataFrame will be applied to the "geometry" column.

22
Measuring area
 To measure the area of each polygon (or MultiPolygon in this
specific case), access the GeoDataFrame.area attribute,
which returns a pandas.Series.
 Note that GeoDataFrame.area is just GeoSeries.area applied to
the active geometry column.

23
Getting polygon boundary and centroid
 To get the boundary of each polygon (LineString), access the
GeoDataFrame.boundary:
 Since we have saved boundary as a new column, we now have
two geometry columns in the same GeoDataFrame.

24
Getting polygon boundary and centroid
 To get the boundary of each polygon (LineString), access the
GeoDataFrame.boundary:
 Since we have saved boundary as a new column, we now have
two geometry columns in the same GeoDataFrame.

25
Measuring distance
 We can also measure how far each centroid is from the first centroid
location.
 Note that geopandas.GeoDataFrame is a subclass of pandas.DataFrame, so
we have all the pandas functionality available to use on the geospatial
dataset — we can even perform data manipulations with the attributes and
geometry information together.

26
Making maps
 GeoPandas can also plot maps, so we can check how the
geometries appear in space.
 To plot the active geometry, call GeoDataFrame.plot().
 To color code by another column, pass in that column as the
first argument.
 In the example below, we plot the active geometry column and
color code by the "area" column. We also want to show a
legend (legend=True).

28
Making maps
 You can also explore your data interactively using
GeoDataFrame.explore(), which behaves in the same way
plot() does but returns an interactive map instead.

29
Making maps
 Switching the active geometry
(GeoDataFrame.set_geometry) to centroids, we can plot the
same data using point geometry..

30
Making maps
 And we can also layer both GeoSeries on top of each other.
We just need to use one plot as an axis for the other.

31
Step 1 : Installing GeoPandas and
Shapely
 need to install the GeoPandas and Shapely libraries in order to
plot a map, and these libraries do not come with the Anaconda
download.
 through the conda-forge channel with the following command
from your terminal
 or alternatively
 Shapely can be pip installed with the command
conda install -c conda-forge geopandas
pip install geopandas
pip install shapely

32
Importing the libraries
 Importing the libraries
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns
import geopandas as gpd
import shapefile as shp
from shapely.geometry import Point
sns.set_style('whitegrid')

33
Load the data
 Use india-polygon.shp shape file to plot India map
 Load the data into a GeoDataFrame as shown below.
fp = r’ india-polygon.shp'
map_df = gpd.read_file(fp)
map_df_copy = gpd.read_file(fp)
map_df.head()

34
Plotting the Shapefiles
 One nice thing about
downloading a
GeoDataFrame, is that
we’ve already got enough
info to make a basic plot.
map_df.plot()

35
Adding better data insights into the
map
df = pd.read_csv('globallandslides.csv')
pd.set_option('display.max_columns', None)df = df[df.country_name=="India"]
df["Year"] = pd.to_datetime(df["event_date"]).dt.year
df = df[df.landslide_category=="landslide"]
state_df = ls_df["admin_division_name"].value_counts()
state_df = state_df.to_frame()
state_df.reset_index(level=0, inplace=True)
state_df.columns = ['State', 'Count']state_df.at[15,"Count"] = 69
state_df.at[0,"State"] = "Jammu and Kashmir" state_df.at[20,"State"] = "Delhi"
state_df.drop(7)

36
 data of the number of
landslides that have happened
in each state over the years

37
Merge the data
 We can now merge this above state data which contains
landslide information with map_df shapefile. We will use the
State column of both the data frames for merging purposes.
#Merging the data
merged = map_df.set_index('st_nm').join(state_df.set_index('State'))
merged['Count'] = merged['Count'].replace(np.nan, 0)
merged.head()

38
Plotting the data on the Shapefile
#Create figure and axes for Matplotlib and set the title
fig, ax = plt.subplots(1, figsize=(10, 10))
ax.axis('off')ax.set_title('Number of landslides in India state-wise',
fontdict={'fontsize': '20', 'fontweight' : '10'})# Plot the figure
merged.plot(column='Count',cmap='YlOrRd', linewidth=0.8, ax=ax,
edgecolor='0',legend=True,markersize=[39.739192, -104.990337],
legend_kwds={'label': "Number of landslides"})

39
Plotting the data on the Shapefile

40
plot the latitudes and longitudes
 We can also plot the latitudes and longitudes
of the occurred landslides on the map along
with the names of states as shown below.
 For this, you will need to find the shapefile
data for Indian states and their latitudes and
longitudes.
 have to plot data points on the map..

�ݺ�ߣ

Geospatial Visualization for datascience

Recommended

More Related Content

Similar to Geospatial Visualization for datascience (20)

More from Sivam Chinna (10)

Recently uploaded (20)

Geospatial Visualization for datascience