Day-03

Pablo Gomez

Review: day 01

x <- seq(from = 5, to = 23, length.out = 10) # create a sequence of numbers
y <- seq(from = 0.1, to = 0.78, length.out = 10) # Create another sequence

mean(x*y) # Get the mean of the multiplication

[1] 7.406667

Objects:
- x
- y

Operators:
- *
- <-

Functions:
- seq()
- mean()

Arguments:
- from
- to
- lengt.out

Review: day 02

select() to select specific columns
slice() to select specific rows based on position
filter() to select specific rows based on a condition
mutate() to create new variables

Review: day 02

Other function we reviewed:

count() Count rows by one or more groups
group_by() aggregate the data by one or more groups
summarise() applies functions to the grouped variables

Review: figures

Components needed to define a graphic:

data, the data set we will use to generate the figure
geometry, or type of graphic we will generate (i.e. histogram, bar, scatter, etc..)
aesthetic or mapping, variables or arguments that will be used for the figure for example: location, color, size, etc..

Review: Maps

Types of vectors:

Point, Location
Line, Location and direction
Polygon, Location and area

Review: Maps

ggplot() +
  geom_stars(data = Mxst) +
  geom_sf(data = Area, fill = NA, col = 'grey60') +
  geom_sf(data = capturesSp, cex = 0.3, col = 'skyblue') +
  theme_void() +
  scale_fill_gradient(low = 'black', high = 'red', na.value = NA) +
  labs(fill = 'Altitude')

Review: Maps

ggplot() + # create the empty canvas
  geom_stars(data = Mxst) +
  geom_sf(data = Area, fill = NA, col = 'grey60') +
  geom_sf(data = capturesSp, cex = 0.3, col = 'skyblue') +
  theme_void() +
  scale_fill_gradient(low = 'black', high = 'red', na.value = NA) +
  labs(fill = 'Altitude')

Review: Maps

ggplot() + # create the empty canvas
  geom_stars(data = Mxst) + # add raster layer
  geom_sf(data = Area, fill = NA, col = 'grey60') + 
  geom_sf(data = capturesSp, cex = 0.3, col = 'skyblue') +
  theme_void() +
  scale_fill_gradient(low = 'black', high = 'red', na.value = NA) +
  labs(fill = 'Altitude')

Review: Maps

ggplot() + # create the empty canvas
  geom_stars(data = Mxst) + # add raster layer
  geom_sf(data = Area, fill = NA, col = 'grey60') + # add polygon layer
  geom_sf(data = capturesSp, cex = 0.3, col = 'skyblue') +
  theme_void() +
  scale_fill_gradient(low = 'black', high = 'red', na.value = NA) +
  labs(fill = 'Altitude')

Review: Maps

ggplot() + # create the empty canvas
  geom_stars(data = Mxst) + # add raster layer
  geom_sf(data = Area, fill = NA, col = 'grey60') + # add polygon layer
  geom_sf(data = capturesSp, cex = 0.3, col = 'skyblue') + # add point layer
  theme_void() +
  scale_fill_gradient(low = 'black', high = 'red', na.value = NA) +
  labs(fill = 'Altitude')

Review: Maps

ggplot() + # create the empty canvas
  geom_stars(data = Mxst) + # add raster layer
  geom_sf(data = Area, fill = NA, col = 'grey60') + # add polygon layer
  geom_sf(data = capturesSp, cex = 0.3, col = 'skyblue') + # add point layer
  theme_void() + # theme for the figure
  scale_fill_gradient(low = 'black', high = 'red', na.value = NA) +
  labs(fill = 'Altitude')

Review: Maps

ggplot() + # create the empty canvas
  geom_stars(data = Mxst) + # add raster layer
  geom_sf(data = Area, fill = NA, col = 'grey60') + # add polygon layer
  geom_sf(data = capturesSp, cex = 0.3, col = 'skyblue') + # add point layer
  theme_void() + # theme for the figure
  scale_fill_gradient(low = 'black', high = 'red', na.value = NA) + # color for the gradient
  labs(fill = 'Altitude')

Review: Maps

ggplot() + # create the empty canvas
  geom_stars(data = Mxst) + # add raster layer
  geom_sf(data = Area, fill = NA, col = 'grey60') + # add polygon layer
  geom_sf(data = capturesSp, cex = 0.3, col = 'skyblue') + # add point layer
  theme_void() + # theme for the figure
  scale_fill_gradient(low = 'black', high = 'red', na.value = NA) + # color for the gradient
  labs(title = 'Map of the study area', fill = 'Altitude') # labels for the figure

Network analysis

Why represent events in a network?

Identify individuals that are very active

Identify individuals that are intermediate

What is a graph?

Elements of a network

Nodes (vertices)

\[V = [1, 2, 3, ..., i]\]

Elements of a network

Nodes (vertices)

\[V = [1, 2, 3, ..., i]\]

Elements of a network

Edges (links)

\[E = [(1, 2), (1, 3), (2, 3), ..., (i,j)]\]

Elements of a network

Network attributes

\[V = [0, 1, 1, ..., x_i]\]

Data structure

General properties

Size
Diameter
Average path length
Density
Fragmentation
Clustering Coefficient

General properties

Local properties

SNA in R

net <- as_tbl_graph(edge)

net

# A tbl_graph: 40 nodes and 1611 edges
#
# A directed multigraph with 1 component
#
# A tibble: 40 × 1
  name 
  <chr>
1 17   
2 12   
3 14   
4 11   
5 7    
6 9    
# ℹ 34 more rows
#
# A tibble: 1,611 × 6
   from    to date    pigs.moved type_orig type_dest
  <int> <int> <chr>        <int> <chr>     <chr>    
1     1     7 8/20/15        160 finisher  sow farm 
2     1     7 8/20/15         76 finisher  sow farm 
3     1     3 9/11/15        155 finisher  nursery  
# ℹ 1,608 more rows

tidygraph

remember %>%?
tidygrapgh introduces two new type of operators:
- %N>% for nodes
- %E>% for edges

Visualization

ggraph(net, layout = 'kk') + # this is our empty canvas
  geom_edge_link(aes(width = pigs.moved)) + # Add the edges
  geom_node_point() + # Add the nodes
  scale_edge_width(range = c(0.01, 0.9)) + # we set the range for the width of the edges
  labs(title = 'Title for the plot', edge_width = 'Number of pigs moved') # labels for the figure

It’s Lab time!

Network analysis in R

Risk assessment

Risk analysis

Risk assessment

Triage

From this we could conclude:

No need for further action
More information needed
Need for a RA

Hazard profiling

A description of the health event and its context, current state of knowledge and potential risk management options

Classification of the causative agent
Persistence in the environment
Pathogenesis
Epidemiology
Prevention and control

Formulating your question

Data collection

To facilitate the search for data, eligibility criteria should be defined that take into account:

Population of interest
Variables of concern
Possible geographical and time restrictions

Expert opinion

Online questionnaire
Expert panel
Interviews
Focus groups discussions

Performing the assessment

Type of outputs might include:

Probability
- Probability of X event happening
- Probability of more than 10 introductions per year
Consequences
- Number of animals/people infected
- Economic impact

Performing the assessment

Risk assessment in R

It’s Lab time!

Risk assessment in R