Network Files¶
Network files are those that define the nodes (wards) of the network, and the links (communiting/movement) of individuals between nodes.
The Network files are a collection of files that must all exist in the same directory. They comprise;
description.json
: This file must exist in the directory, and provides the metadata needed to locate all of the other necessary files.work_size / play_size
: These files list the index and population ofworkers (in
work_size
) and players (inplay_size
) in each ward in the network. The total population in each ward (node) is the sum of the work and play populations.
work
: This file lists all of the work connections between nodes, giving the population of workers who commute from the source node to the destination.play
: This file lists all of the play connections between nodes, giving the proportion of players who make random movements from the source node to the destination node.position
: This file contains the coordinates of the centre of each node. These coordinates can be X/Y or latitude/longitude coordinates.lookup
: This file contains the names and other metadata about each node, e.g. the name, local authority, region etc.
These files are all described below.
description.json¶
This is a JSON-formatted file that provides metadata about the network model that is contained in the directory. The file describes a simple dictionary of the following key-value pairs;
work
: Filename of thework
file.work_size
: Filename of thework_size
file.play
: Filename of theplay
file.play_size
: Filename of theplay_size
file.position
: Filename of theposition
file.coordinates
: Whether the coordinates in theposition
file are in x/y (x/y
) or latitude/longitude (lat/long
) format.coordinate_units
: (Optional) - the units for x/y coordinates. This should be eitherm
for meters, orkm
for kilometers. Distances inmetawards
are always reported as kilometers.lookup
: (Optional) Filename of thelookup
file.lookup_columns
: (Optional) A dictionary that gives the column numbers (zero-indexed)for thecode
,name
,alternate_code
,alternate_name
,authority_code
andauthority_name
fields for the ward metadata.
work¶
The work
file contains the list of all work connections between nodes,
giving the number of individuals who commute from the source node to the
destination node. This is a column-based file with three columns of numbers
that can be space or comma separated. For example;
1 1 290
1 2 3
1 5 139
1 6 59
1 7 17
1 8 119
1 9 37
1 10 121
The first number is the (1-indexed) index of the source node, while the second number is the (1-indexed) index of the destination node. The third number is the number of individuals who commute from the source to the destination node.
In this example, this lists workers who commute from ward 1 to wards 1 to 10. 290 workers commute from ward 1 to ward 1 (so work in the same ward in which they live). 3 workers commute from ward 1 to ward 2, 139 commute from ward 1 to ward 5 etc.
Note
All of the source wards have to be listed contiguously, i.e.
you must list all of the connections where the source ward is
equal to 1
before you list all of the connections where the
source ward is equal to 2
etc.
work_size¶
The work_size
file contains the number of workers who reside in each ward.
This is a column-based file with two columns of numbers that can be space or
comma separated. For example;
1 6800
2 1091
3 7148
4 5684
5 7226
6 6561
7 6904
8 7213
9 6715
10 7452
The first number is the (1-indexed) index of the ward, while the second
is the number of workers in that ward. The number of workers in a ward must
equal the sum of the number of workers from the work
file that say
that they commute from that ward.
In this example, ward 1 has 6800 workers, ward 2 has 1091 workers etc.
play¶
The play
file contains the list of play connections between wards.
This is a column-based file with three columns of numbers that can be
space or comma separated. For example;
1 1 0.0426470588235294
1 2 0.000441176470588235
1 5 0.0204411764705882
1 6 0.00867647058823529
1 7 0.0025
1 8 0.0175
1 9 0.00544117647058823
1 10 0.0177941176470588
The first number is the (1-indexed) index of the source ward, while the second is the (1-indexed) index of the destination ward. The third number is the fraction of players who will randomly travel from the source ward to the destination ward.
In this case, 0.0426...
of players will remain in ward 1, while
0.00044...
of players will randomly move from ward 1 to ward 2.
Note
Be careful as the third number is a fraction (floating point number
between 0 and 1) and not the number of players that move. Also note
that the same ordering requirements as for the work
file apply,
namely that all connections for source ward 1 have to be listed before
all connections for source ward 2 etc.
play_size¶
The play_size
file contains the number of players in each ward.
This is a column-based file with two columns of numbers that can be
space or comma separated. For example:
1 8915
2 374
3 7012
4 10579
5 8703
6 12257
7 10533
8 11259
9 8592
10 10999
The first number is the (1-indexed) index of the ward, while the second is the number of players in that ward. In this case, there are 8915 players in ward 1, 374 players in ward 2 etc.
position¶
The position
file contains the coordinates of the center of each ward.
This is a column-based file with three columns of numbers that can be
space or comma separated. For example:
1 524693.890435782 190136.324582048
2 532169.852194767 181663.72329877
3 522106.698233411 179737.792519091
4 533388.404693453 193451.026467071
5 525973.729674732 188464.548078951
6 523953.505171282 187729.710513154
7 520763.019103366 193360.422742775
8 523100.570665414 189524.60815157
9 523747.496973864 196638.656771657
10 523019.549069799 192050.596775831
The coordinates are either x/y or latitude/longitude, depending on what
is set in the description.json
file. If x/y, then the units are also
set in this file.
In this case, this is for x/y coordinates in meters. The first column is the (1-indexed) index of the ward. The second column is the X coordinate, while the third is the Y coordinate.
Note
If lat/long was used, then the second column would be the latitude, and the third column the longitude.
So ward 1 has its center are (524.7km, 190.1km), while ward 2 has its center at (532.1km, 181.7km) etc.
lookup¶
The lookup
file contains metadata information about a ward that allows
it to be looked up by name, region etc. This is a column-based file with
a header row, and columns separated by spaces or commas. For example:
WD11CD,WD11NM,WD11NMW,CMWD11CD,CMWD11NM,CMWD11NMW,IND,LAD11CD,LAD11NM,LAD11NMW,FID
E05002337,Central,,E36000890,Central,,0,E06000039,Slough,,2001
E05002338,Chalvey,,E36000891,Chalvey,,0,E06000039,Slough,,2002
E05002339,Cippenham Green,,E36000892,Cippenham Green,,0,E06000039,Slough,,2003
E05002340,Cippenham Meadows,,E36000893,Cippenham Meadows,,0,E06000039,Slough,,2004
E05002341,Colnbrook with Poyle,,E36000894,Colnbrook with Poyle,,0,E06000039,Slough,,2005
E05002342,Farnham,,E36000895,Farnham,,0,E06000039,Slough,,2006
E05002343,Foxborough,,E36000896,Foxborough,,0,E06000039,Slough,,2007
E05002344,Haymill,,E36000897,Haymill,,0,E06000039,Slough,,2008
E05002345,Kedermister,,E36000898,Kedermister,,0,E06000039,Slough,,2009
E05002346,Langley St Mary's,,E36000899,Langley St Mary's,,0,E06000039,Slough,,2010
The first row provides the names of each of the columns. These are ignored by the code. Each row gives the information for the ward whose index is equal to the line number (so line 1 give the information for ward at index 1).
The lookup_columns
field in description.json
specifies which columns
from this file to use for the name, code, authority etc. for each ward.
For example, this has set column 0 for the code, column 1 for the name,
and so the first ward has code E05002337
and name Central
.