Machine Learning On Geographical Data Using Python Introduction Into Geodata With Applications And Use Cases Joos Korstanje instant download
Machine Learning On Geographical Data Using Python Introduction Into Geodata With Applications And Use Cases Joos Korstanje instant download
https://ebookbell.com/product/machine-learning-on-commodity-tiny-
devices-theory-and-practice-song-guo-47219342
https://ebookbell.com/product/machine-learning-on-kubernetes-a-
practical-handbook-for-building-and-using-a-complete-open-source-
machine-learning-platform-on-kubernetes-masood-56267644
https://ebookbell.com/product/practical-machine-learning-on-
databricks-seamlessly-transition-ml-models-and-mlops-on-databricks-
debu-sinha-53889210
https://ebookbell.com/product/automated-machine-learning-on-aws-
fasttrack-the-development-of-your-productionready-machine-learning-
applications-the-aws-way-1st-edition-trenton-potgieter-42751874
Mastering Machine Learning On Aws Advanced Machine Learning In Python
Using Sagemaker Apache Spark And Tensorflow Dr Saket Sr Mengle
https://ebookbell.com/product/mastering-machine-learning-on-aws-
advanced-machine-learning-in-python-using-sagemaker-apache-spark-and-
tensorflow-dr-saket-sr-mengle-11117496
https://ebookbell.com/product/practical-machine-learning-on-
databricks-seamlessly-transition-ml-models-and-mlops-on-databricks-
anonymous-55324496
https://ebookbell.com/product/practical-machine-learning-on-
databricks-1st-edition-debu-sinha-56235472
https://ebookbell.com/product/mastering-machine-learning-on-aws-saket-
s-r-mengle-170428468
https://ebookbell.com/product/handson-machine-learning-on-google-
cloud-platform-giuseppe-ciaburro-48958816
Machine Learning
on Geographical
Data Using Python
Introduction into Geodata with
Applications and Use Cases
—
Joos Korstanje
Machine Learning
on Geographical
Data Using Python
Introduction into Geodata
with Applications and Use Cases
Joos Korstanje
Machine Learning on Geographical Data Using Python: Introduction into Geodata
with Applications and Use Cases
Joos Korstanje
VIELS MAISONS, France
Introduction�������������������������������������������������������������������������������������������������������������xv
iii
Table of Contents
iv
Table of Contents
v
Table of Contents
vi
Table of Contents
vii
Table of Contents
viii
Table of Contents
Index��������������������������������������������������������������������������������������������������������������������� 307
ix
About the Author
Joos Korstanje is a data scientist, with over five years
of industry experience in developing machine learning
tools. He has a double MSc in Applied Data Science and
in Environmental Science and has extensive experience
working with geodata use cases. He has worked at a
number of large companies in the Netherlands and France,
developing machine learning for a variety of tools. His
experience in writing and teaching has motivated him
to write this book on machine learning for geodata with
Python.
xi
About the Technical Reviewer
Xiaochi Liu is a PhD researcher and data scientist at
Macquarie University, specializing in machine learning,
explainable artificial intelligence, spatial analysis, and their
novel application in environmental and public health. He is
a programming enthusiast using Python and R to conduct
end-to-end data analysis. His current research applies
cutting-edge AI technologies to untangle the causal nexus
between trace metal contamination and human health
to develop evidence-based intervention strategies for
mitigating environmental exposure.
xiii
Introduction
Spatial data has long been an ignored data type in general data science and statistics
courses. Yet at the same time, there is a field of spatial analysis which is strongly
developed. Due to differences in tools and approaches, the two fields have long
developed in separate environments.
With the popularity of data in many business environments, the importance of
treating spatial data is also increasing. The goal of the current book is to bridge the gap
between data science and spatial analysis by covering tools of both worlds and showing
how to use tools from both to answer use cases.
The book starts with a general introduction to geographical data, including data
storage formats, data types, common tools and libraries in Python, and the like. Strong
attention is paid to the specificities of spatial data, including coordinate systems
and more.
The second part of the book covers a number of methods of the field of spatial
analysis. All of this is done in Python. Even though Python is not the most common
tool in spatial analysis, the ecosystem has taken large steps in user-friendliness and has
great interoperability with machine learning libraries. Python with its rich ecosystem of
libraries will be an important tool for spatial analysis in the near future.
The third part of the book covers multiple machine learning use cases on spatial
data. In this part of the book, you see that tools from spatial analysis are combined with
tools from machine learning and data science to realize more advanced use cases than
would be possible in many spatial analysis tools. Specific considerations are needed for
applying machine learning to spatial data, due to the specific nature of coordinates and
other specific data formats of spatial data.
Source Code
All source code used in the book can be downloaded from g ithub.com/apress/
machine-learning-geographic-data-python.
xv
PART I
General Introduction
CHAPTER 1
Introduction to Geodata
Mapmaking and analysis of the geographical environment around us have been present
in nature and human society for a long time. Human maps are well known to all of us:
they are a great way to share information about our environment with others.
Yet communicating geographical instructions is not invented only by the human
species. Bees, for example, are well known to communicate on food sources with their
fellow hive mates. Bees do not make maps, but, just like us, they use a clearly defined
communication system.
As geodata is the topic of this book, I find it interesting to share this out-of-the-box
geodata system used by honeybees. Geodata in the bee world has two components:
distance and direction.
Honeybee distance metrics
–– The round dance: A food source is present less than 50 meters from
the hive.
–– The sickle dance: Food sources are present between 50 and 150
meters from the hive.
–– The waggle (a.k.a. wag-tail) dance: Food sources are over 150 meters
from the hive. In addition, the duration of the waggle dance is an
indicator of how far over 150 meters the source is located.
–– As the sun changes location throughout the day, bees will update
each other by adapting their communication dances accordingly.
3
© Joos Korstanje 2022
J. Korstanje, Machine Learning on Geographical Data Using Python,
https://doi.org/10.1007/978-1-4842-8287-8_1
Chapter 1 Introduction to Geodata
Geodata Definitions
To get started, I want to cover the basics of coordinate systems in the simplest
mathematic situation: the Euclidean space. Although the world does not respect
the hypothesis made by Euclidean geometry, it is a great entry into the deeper
understanding of coordinate systems.
4
Chapter 1 Introduction to Geodata
Cartesian Coordinates
To locate points in the Euclidean space, we can use the Cartesian coordinate system.
This coordinate system specifies each point uniquely by a pair of numerical coordinates.
For example, look at the coordinate system in Figure 1-2, in which two points are located:
a square and a triangle.
The square is located at x = 2 and y = 1 (horizontal axis). The triangle is located
at x = -2 and y = -1.
5
Chapter 1 Introduction to Geodata
The point where the x and y axes meet is called the Origin, and distances are
measured from there. Cartesian coordinates are among the most well-known coordinate
system and work easily and intuitively in the Euclidean space.
In this schematic drawing, the star is designated as the pole, and the thick black line
to the right is chosen as the polar axis. This system is quite different from the Cartesian
system but still allows us to identify the exact same points: just in a different way.
The points are identified by two components: an angle with respect to the polar axis
and a distance. The square that used to be referred to as Cartesian coordinate (2,1) can
be referred to by an angle from the polar axis and a distance.
6
Chapter 1 Introduction to Geodata
At this point, you can measure the distance and the angle and obtain the coordinate
in the polar system. Judged by the eye alone, we could say that the angle is probably
more or less 30° and the distance is slightly above 2. We would need to have more precise
measurement tools and a more precise drawing for more precision.
There are trigonometric computations that we can use to convert between polar and
Cartesian coordinates. The first set of formulas allows you to go from polar to Cartesian:
The letter r signifies the distance and the letter φ is the angle. You can go the other
way as well, using the following formulas:
As a last part to cover about degrees, I want to mention the equivalence between
measuring angles in degrees and in radians. The radian system may seem scary if you
are not used to it, but just remember that for every possible angle that you can measure
(from 0 to 360) there is a corresponding notation in the radian system. Figure 1-5
shows this.
7
Chapter 1 Introduction to Geodata
8
Chapter 1 Introduction to Geodata
ArcGIS
ArcGIS, made by ESRI, is arguably the most famous software package for working with
Geographic Information Systems. It has a very large number of functionalities that can
be accessed through a user-friendly click-button system, but visual programming of
geodata processing pipelines is also allowed. Python integration is even possible for
those who have specific tasks for which there are no preexisting tools in ArcGIS. Among
its tools are also AI and data science options.
ArcGIS is a great software for working with geodata. Yet there is one big
disadvantage, and that is that it is a paid, proprietary software. It is therefore accessible
only to companies or individuals that have no difficulty paying the considerably high
price. Even though it may be worth its price, you’ll need to be able to pay or convince
your company to pay for such software. Unfortunately, this is often not the case.
This approach can be a good fit for your need if you are not afraid to commit to a
system like QGIS and fill the gaps that you may eventually encounter.
Python/R Programming
Finally, you can use Python or R programming for working with geodata as well.
Programming, especially in Python or R, is a very common skill among data
professionals nowadays.
As programming skills were less well spread a few years back, the boom in data
science, machine learning, and artificial intelligence has made languages like Python
become very commonly spread throughout the workforce.
Now that many are able to code or have access to courses to learn how to code, the
need for full software becomes less. The availability of a number of well-functioning
geodata packages is enough for many to get started.
Python or R programming is a great tool for treating geodata with common or more
modern methods. By using these programming languages, you can easily apply tools
from other libraries to your geodata, without having to convert this to QGIS modules, for
example.
The only problem that is not very well solved by programming languages is long-
term geodata storage. For this, you will need a database. Cloud-based databases are
nowadays relatively easy to arrange and manage, and this problem is therefore relatively
easily solved.
10
Chapter 1 Introduction to Geodata
Shapefile
The shapefile is a very commonly used file format for geodata because it is the standard
format for ArcGIS. The shapefile is not very friendly for being used outside of ArcGIS, but
due to the popularity of ArcGIS, you will likely encounter shapefiles at some point.
The shapefile is not really a single file. It is actually a collection of files that are
stored together in one and the same directory, all having the same name. You have the
following files that make up a shapefile:
As an example, let’s look at an open data dataset containing the municipalities of the
Paris region that is provided by the French government. This dataset is freely available
at https://geo.data.gouv.fr/en/datasets/8fadd7040c4b94f2c318a0971e8fae
db7b5675d6
On this website, you can download the data in SHP/L93 format, and this will allow
you to download a directory with a zip file. Figure 1-6 shows what this contains.
As you can see, there are the .shp file (the main file), the .shx file (the index file), the
.dbf file containing the attributes, and finally the optional .prj file.
11
Chapter 1 Introduction to Geodata
For this exercise, if you want to follow along, you can use your local environment or a
Google Colab Notebook at https://colab.research.google.com/.
You have to make sure that in your environment, you install geopandas:
Then, make sure that in your environment you have a directory called Communes_
MGP.shp in which you have the four files:
–– Communes_MGP.shp
–– Communes_MGP.dbf
–– Communes_MGP.prj
–– Communes_MGP.shx
In a local environment, you need to put the “sample_data” file in the same directory
as the notebook, but when you are working on Colab, you will need to upload the whole
folder to your working environment, by clicking the folder icon and then dragging and
dropping the whole folder onto there. You can then execute the Python code in Code
Block 1-1 to have a peek inside the data.
12
Chapter 1 Introduction to Geodata
To make something more visual, you can use the code in Code Block 1-2.
shapefile.plot()
13
Chapter 1 Introduction to Geodata
You will obtain the map corresponding to this dataset as in Figure 1-8.
Figure 1-8. The map resulting from Code Block 1-2. Image by author
Data source: Ministry of DINSIC. Original data downloaded from https://geo.
data.gouv.fr/en/datasets/8fadd7040c4b94f2c318a0971e8faedb7b5675d6,
updated on 1 July 2016. Open Licence 2.0 (www.etalab.gouv.fr/wp-content/
uploads/2018/11/open-licence.pdf)
14
Chapter 1 Introduction to Geodata
If you try opening the file with a text editor, you’ll find that it is an XML file (very
summarized, XML is a data storage pattern that can be recognized by many < and > signs).
Compared to the shapefile, you can see that KML is much easier to understand and
to parse. A part of the file contents is shown in Figure 1-9.
To get a KML file into Python, we can again use geopandas. This time, however, it is
a bit less straightforward. You’ll also need the Fiona package to obtain a KML driver. The
total code is shown in Code Block 1-3.
import fiona
gpd.io.file.fiona.drvsupport.supported_drivers['KML'] = 'rw'
You’ll then see the exact same geodataframe as before, which is shown in
Figure 1-10.
As before, you can plot this geodataframe to obtain a basic map containing the
municipalities of the area of Paris and around. This is done in Code Block 1-4.
16
Chapter 1 Introduction to Geodata
kmlfile.plot()
Figure 1-11. The plot resulting from Code Block 1-4. Screenshot by author
Data source: Ministry of DINSIC. Original data downloaded from https://geo.
data.gouv.fr/en/datasets/8fadd7040c4b94f2c318a0971e8faedb7b5675d6,
updated on 1 July 2016. Open Licence 2.0 (www.etalab.gouv.fr/wp-content/
uploads/2018/11/open-licence.pdf)
An interesting point here is that the coordinates do not correspond with the map that
was generated from the shapefile. If you’ve read the first part of this chapter, you may
have a hinge on how this is caused by coordinate systems. We’ll get into this in much
more detail in Chapter 2.
GeoJSON
The json format is a data format that is well known and loved by developers. Json is
much used in communication between different information systems, for example, in
website and Internet communication.
The json format is loved because it is very easy to parse, and this makes it a perfect
storage for open source and other developer-oriented tools.
17
Chapter 1 Introduction to Geodata
Json is a key-value dataset, which is much like the dictionary in Python. The whole is
surrounded by accolades. As an example, I could write myself as a json object as in this
example:
{ 'first_name': 'joos',
'last_name': 'korstanje',
'job': 'data scientist' }
As you can see, this is a very flexible format, and it is very easy to adapt to all kinds of
circumstances. You might easily add GPS coordinates like this:
{ 'first_name': 'joos',
'last_name': 'korstanje',
'job': 'data scientist',
'latitude': '48.8566° N',
'longitude': '2.3522° E' }
18
Chapter 1 Introduction to Geodata
You can get a GeoJSON file easily into the geopandas library using the code in Code
Block 1-5.
As expected, the data looks exactly like before (Figure 1-13). This is because it is
transformed into a geodataframe, and therefore the original representation as json is not
maintained anymore.
19
Chapter 1 Introduction to Geodata
You can make the plot of this geodataframe to obtain a map, using the code in Code
Block 1-6.
geojsonfile.plot()
20
Chapter 1 Introduction to Geodata
Figure 1-14. The plot resulting from Code Block 1-6. Image by author
Data source: Ministry of DINSIC. Original data downloaded from https://geo.
data.gouv.fr/en/datasets/8fadd7040c4b94f2c318a0971e8faedb7b5675d6,
updated on 1 July 2016. Open Licence 2.0 (www.etalab.gouv.fr/wp-content/
uploads/2018/11/open-licence.pdf)
TIFF/JPEG/PNG
Image file types can also be used to store geodata. After all, many maps are 2D images
that lend themselves well to be stored as an image. Some of the standard formats to store
images are TIFF, JPEG, and PNG.
–– The PNG format is another well-known image file format. You can
make this file into a GeoJPEG as well when using it together with a
PWG (world file).
21
Chapter 1 Introduction to Geodata
Image file types are generally used to store raster data. For now, consider that raster
data is image-like (one value per pixel), whereas vector data contains objects like lines,
points, and polygons. We’ll get to the differences between raster and vector data in a
next chapter.
On the following website, you can download a GeoTIFF file that contains an
interpolated terrain model of Kerbernez in France:
https://geo.data.gouv.fr/en/datasets/b0a420b9e003d45aaf0670446f0d600df
14430cb
You can use the code in Code Block 1-7 to read and show the raster file in Python.
Note Depending on your OS, you may obtain a .tiff file format rather than a
.tif when downloading the data. In this case, you can simply change the path to
become .tiff, and the result should be the same. In both cases, you will obtain the
image shown in Figure 1-15.
22
Chapter 1 Introduction to Geodata
Figure 1-15. The plot resulting from Code Block 1-7. Image by author
Data source: Ministry of DINSIC. Original data downloaded from https://geo.
data.gouv.fr/en/datasets/b0a420b9e003d45aaf0670446f0d600df14430cb,
updated on “unknown.” Open Licence 2.0 (www.etalab.gouv.fr/wp-content/
uploads/2018/11/open-licence.pdf)
It is interesting to look at the coordinates and observe that this file’s coordinate
values are relatively close to the first file.
CSV/TXT/Excel
The same file as used in the first three examples is also available in CSV. When
downloading it and opening it with a text viewer, you will observe something like
Figure 1-16.
23
Chapter 1 Introduction to Geodata
The important thing to take away from this part of the chapter is that geodata is “just
data,” but with geographic references. These can be stored in different formats or in
different coordinate systems to make things complicated, but in the end you must simply
make sure that you have some sort of understanding of what you have in your data.
You can use many different tools for working with geodata. The goal of those tools
is generally to make your life easier. As a last step for this introduction, let’s have a short
introduction to the different Python tools that you may encounter on your geodata
journey.
24
Chapter 1 Introduction to Geodata
Key Takeaways
1. Cartesian coordinates and polar coordinates are two alternative
coordinate systems that can indicate points in a two-dimensional
Euclidean space.
26
CHAPTER 2
Coordinate Systems
and Projections
In the previous chapter, you have seen an introduction to coordinate systems. You saw
an example of how you can use Cartesian coordinates as well as polar coordinates to
identify points on a flat, two-dimensional Euclidean space. It was already mentioned at
that point that the real-world scenario is much more complex.
When you are making maps, you are showing things (objects, images, etc.) that are
located on earth. Earth does not respect the rules that were shown in the Euclidean
example because Earth is an ellipsoid: a ball form that is not perfectly round. This makes
map and coordinate system calculations much more complex than what high-school
mathematics teaches us about coordinates.
To make the problem clearer, let’s look at an example of airplane navigation.
Airplane flights are a great example to illustrate the problem, as they generally cover
long distances. Taking into account the curvature of the earth really doesn’t matter much
when measuring the size of your terrace, but it does make a big impact when moving
across continents.
Imagine you are flying from Paris to New York using this basic sketch of the world’s
geography. You are probably well aware of such an organization of the world’s map on a
two-dimensional image.
A logical first impression would be that to go from Madrid to New York in the
quickest way, we should follow a line parallel from the latitude lines. Yet (maybe
surprisingly at first) this is not the shortest path. An airplane would better curve via
the north!
The reason for this is that the more you move to the north, the shorter the latitude
lines actually are. Latitude lines go around the earth, so at the North Pole you have a
length of zero, and at the equator, the middle line is the longest possible. The closer to
the poles, the shorter the distance to go around the earth.
27
© Joos Korstanje 2022
J. Korstanje, Machine Learning on Geographical Data Using Python,
https://doi.org/10.1007/978-1-4842-8287-8_2
Chapter 2 Coordinate Systems and Projections
As this example takes place in the northern hemisphere, the closest pole is the North
Pole. By curving north on the northern hemisphere (toward the pole), an airplane can
get to its destination with fewer kilometers. Figure 2-1 illustrates this.
Let’s now consider an example where you are holding a round soccer ball. When
going from one point to another on a ball, you will intuitively be able to say which path is
the fastest. If you are looking straight at the ball, when following your finger going from
one point to another, you will see your hand making a shape like in Figure 2-2.
28
Chapter 2 Coordinate Systems and Projections
Figure 2-2. The shortest path on a ball is not a straight line in two-dimensional
view. Image by author
When making maps, we cannot plot in three dimensions, and we, therefore,
need to find some way or another to put a three-dimensional path onto a two-
dimensional image.
Many map makers have proposed all sorts of ways to solve this unsolvable problem,
and the goal of this chapter is to help you understand how to deal effectively with those
3D to 2D mapping distortions that will be continuously looking to complexify your work
on geodata.
Coordinate Systems
While the former discussion was merely intuitive, it is now time to slowly get to more
official definitions of the concepts that you have seen. As we are ignoring the height of
a point (e.g., with respect to sea level) for the moment, we can identify three types of
coordinate systems:
29
Chapter 2 Coordinate Systems and Projections
30
Chapter 2 Coordinate Systems and Projections
There are more specific definitions that define the WGS 84, yet at this point, the
information becomes very technical. To quote from the Wikipedia page of the WGS 84:
The WGS 84 datum surface is an oblate spheroid with equatorial radius a
= 6378137 m at the equator and flattening f = 1/298.257223563. The refined
value of the WGS 84 gravitational constant (mass of Earth's atmosphere
included) is GM = 3986004.418×108 m3/s2. The angular velocity of the
Earth is defined to be ω = 72.92115×10−6 rad/s.
You are absolutely not required to memorize any of those details. I do hope that it
gives you an insight into how detailed a definition of Geographic Coordinate Systems
has to be. This explains how it is possible that other people and organizations have
identified alternate definitions. This is why there are many coordinate systems out there
and also one of the reasons why working with geospatial data can be hard to grasp in the
beginning.
31
Chapter 2 Coordinate Systems and Projections
X and Y Coordinates
When working with Projected Coordinate Systems, we do not talk about latitude and
longitude anymore. As latitude and longitude are relevant only for measurements
on the globe (ellipsoid), but on a flat surface, we can drop this complexity. Once the
three-dimensional lat/long coordinates have been converted to the coordinates of their
projection, we simply talk about x and y coordinates.
X is generally the distance to the east starting from the origin and y the distance to
the north starting from the origin. The location of the origin depends on the projection
that you are using. The measurement unit also changes from one Projected Coordinate
System to another.
32
Chapter 2 Coordinate Systems and Projections
The Albers equal area conic projection takes a very different approach, as it is conic.
Making conic maps is often done to make some zones better represented. The Albers
equal area conic projection, also called the Albert projection, projects the world to a two-
dimensional map while respecting areas as shown in Figure 2-4.
33
Chapter 2 Coordinate Systems and Projections
Figure 2-4. The world seen in an Albers equal area conic projection
Source: https://commons.wikimedia.org/wiki/File:World_borders_
albers.png. Public Domain
Conformal Projections
If shapes are important for your use case, you may want to use a conformal projection.
Conformal projections are designed to preserve shapes. At the cost of distorting the areas
on your map, this category of projections guarantees that all of the angles are preserved,
and this makes sure that you see the “real” shapes on the map.
Mercator
The Mercator map is very well known, and it is the standard map projection for many
projects. Its advantage is that it has north on top and south on the bottom while
preserving local directions and shapes.
Unfortunately, locations far away from the equator are strongly inflated, for example,
Greenland and Antarctica, while zones on the equator look too small in comparison
(e.g., Africa).
34
Chapter 2 Coordinate Systems and Projections
The Lambert conformal conic projection is another conformal projection, meaning that
it also respects local shapes. This projection is less widespread because it is a conic map,
and conic maps have never become as popular as rectangles. However, it does just as
well on plotting the earth while preserving shapes, and it has fewer problems with size
distortion. It looks as shown in Figure 2-6.
35
Chapter 2 Coordinate Systems and Projections
Equidistant Projections
As the name indicates, you should use equidistant projections if you want a map
that respects distances. In the two previously discussed projection types, there is no
guarantee that distance between two points is respected. As you can imagine, this will be
a problem for many use cases. Equidistant projections are there to save you if distances
are key to your solution.
36
Chapter 2 Coordinate Systems and Projections
The equidistant conic projection is another conic projection, but this time it preserves
distance. It is also known as the simple conic projection, and it looks as shown in
Figure 2-8.
37
Chapter 2 Coordinate Systems and Projections
One example of an azimuthal projection is the Lambert equal area azimuthal. As the
name indicates, it is not just azimuthal but also equal area. The world according to this
projection looks as shown in Figure 2-9.
38
Another Random Document on
Scribd Without Any Related Topics
pareillement d'eulx, il fût uzé de quelque respect pour l'amour de
vous.
A quoy, pour le regard de la dicte Royne d'Escoce, ilz ne m'ont
donné meilleure satisfaction que de m'alléguer plusieurs occasions
d'offance que la Royne d'Angleterre prétend contre elle, et qu'on
vous fera une telle déclaration de ce qu'elle avoit projetté de faire,
pour se soustraire de vostre alliance, que vous n'aurez plus ocasion
d'avoir soing, ny souvenance d'elle; et, au regard des Escouçoys, ilz
m'ont respondu qu'ilz feront en sorte que la Royne, leur Mestresse, y
procèdera, le plus qu'il sera possible, sellon vostre desir et intention;
et sur le reste de la négociation que j'ay continué avec eulx, despuys
ma dernière audience, ilz m'ont résoluement asseuré que la dicte
Dame envoyera bientost ung principal seigneur de ce conseil devers
Vostre Majesté. Et je pense avoir desjà tant faict, Sire, que ce sera
milord de Burgley, mais quant j'en seray encores plus certain, et que
je sçauray le temps de son partement, j'en advertiray en dilligence
Vostre Majesté, ayant opinion que de son voyage et de ceste sienne
commission a de résulter tout l'effect de ce que pouvez espérer de
ceste princesse et de ce royaulme. Sur ce, etc.
e
Ce xv jour d'octobre 1571.
A la Royne.
Madame, j'ay fort curieusement considéré les propos qui ont esté
tenuz, entre Vostre Majesté et l'ambassadeur d'Angleterre, sellon
qu'ilz sont fort bien et fort dilligentement recueilliz, en la lettre qu'il
vous a pleu m'escripre du xxviiie du passé. Et, pour le regard de ce
qu'il a commancé de vous faire quelque pleinte de moy, je sçay,
Madame, que je vous ay ordinairement randu ung si véritable et si
particullier compte, de tout ce que j'ay dict et négocié par deçà, qu'il
ne vous a peu dire rien de nouveau, aussi ne veulx je faillyr de
remercyer très humblement Vostre Majesté pour la favorable
responce que luy avez faicte de la bonne opinion, en quoy il playt au
Roy et à vous me tenir, laquelle me suffit pour l'entière justiffication
de mes actions, qui ne sont vouez qu'au seul service de Voz
Majestez; et j'espère, Madame, que, dans peu de jours, vous
l'ouyrez parler en aultre façon de moy, sellon que la Royne, sa
Mestresse, et ses deux principaulx conseillers m'ont dict, touchant
l'inquisition qu'ilz avoient faicte de moy à cause de ces deux mil
escuz, qu'il n'a esté trouvé que j'aye jamais faict ny dict chose, en
ceste charge, qui ne soit bonne et honneste. Il est vray, Madame,
que j'eusse bien vollu qu'il vous fût souvenu de luy parler du dict
argent en la façon que auparavant j'en avois escript, mais cella se
pourra rabiller la première foys que luy donrez audience, et suys très
ayse que luy ayez ainsy sagement et vertueusement respondu,
comme avez faict, touchant la Royne d'Escoce, affin qu'en la
manière de procéder, dont l'on use icy contre elle et contre les
Escouçoys, l'on y aille plus réservé. Et quant au propos du mariage,
si j'eusse heu vostre lettre avant aller à l'audience, j'eusse suyvy
exactement les termes d'icelle, tant y a que je n'ay point outrepassé
ceulx de la précédante dépesche du xxviie: et est à considérer,
Madame, qu'en telles matières, il se trouve toutjour d'honnestes
excuses et interruptions jusques à la porte de l'esglize. Je crains
seulement que ceste expression: «de vouloir avoir l'exercice public et
libre de la religion,» si le Sr de Valsingam en escript par deçà, ne
réfroydisse ou ceste Royne d'envoyer devers Voz Majestez, ou milord
de Burgley de faire le voyage; tant y a que j'en mèneray la pratique
ainsy soubdain et chauldement comme je l'ay commancée. Et, au
regard d'introduyre le segond propos de mariage, il semble,
Madame, qu'il sera beaucoup meilleur d'atandre à le toucher sur
quelque occasion des choses que milord de Burgley pourra dire ou
proposer par dellà, car je voys bien qu'il n'est encores temps d'en
parler icy; tant y a que, en ceste et aultres particullaritez de vostre
lettre, je métray peyne d'y observer le temps et l'ocasion pour m'y
conduyre tout ainsy qu'il vous playt me le commander. Sur ce, etc.
Ce xve jour d'octobre 1571.
CCXIIe DÉPESCHE
e
—du xx jour d'octobre 1571.—
(Envoyée exprès jusques à Calais par Jehan Monyer.)
Affaires d'Écosse.—Assurance donnée par Burleigh qu'Élisabeth a renoncé à
user de rigueur contre Marie Stuart, et que tout envoi de secours en
Écosse est suspendu.—Procédure contre le duc de Norfolk.—Arrestation
de lord Coban.—Fuite du comte Derby.
Au Roy.
Au Roy.
Au Roy.
Au Roy.
Au Roy.
A la Royne.
Au Roy.
ebookbell.com