Extra seeds files¶
This file specifies how infections should be seeded in an outbreak. This is a column-based flexible-format file, with each line of the file containing information about a specific seeding event.
There should be up to three pieces of information per line;
day
: The day or date to seed the infection(s). This should be an integer or a date that is interpreted viaInterpret.day_or_date
function. This either sets the day number on which to seed, or the exact date in which to seed.ward
: The ward (either index or name) in which the infection will be seeded. This should either be an integer (whole number) specifying the index of the ward (1-indexed), or a string that identifies the ward by name, e.g.Clifton/Bristol
would look for the ward calledClifton
in the authorityBristol
. The ward is looked up by name using theWardInfos.find
function.number
: The number of infections to be seeded. This is an integer that is interpreted via theInterpret.integer
function.demographic
(optional) : The demographic (either index or name) in which the infection will be seeded. If this is not specified, then the first demographic in the network is seeded. If this is the index it is interpreted viaInterpret.integer
, while if this is the name then this will direct look-up the demographic by name.
This is a column-formatted file. Like all flexible-format files you can use
commas or spaces to separate columns. The separators used in the first line
will be assumed to be used for the rest of the file. You can order the
columns however you wish, as long as you provide column headers
(day
, ward
, number
, and (optionally) demographic
). Otherwise
the column orders are day
, number
, ward
,
(optional) demographic
.
You can add extra spaces and blank lines to make the file more readable,
and can add comments via the #
character.
Examples¶
1 5 1
Seed 5 infections on day 1 (first day) of the outbreak in ward 1.
1, 5, 1
Same, but using commas to separate
day, ward, number
# seed 5 infections on day 1 in ward 1
1, 5, 1
Same, but naming the columns and adding a comment
# Seeding two locations in Bristol
day, demographic, ward, number
tomorrow, 0, Clifton / Bristol, "rand(5, 20)"
next week, 0, Knowle / Bristol, 5 + 3
Adding comments and changing the order of columns. Seeding by fuzzy dates,
e.g. tomorrow
and next week
in wards identified by
name / authority
. Number to be seeded is a random number from 5 to 20
(inclusive) in Clifton, Bristol, and the result of the expression
5 + 3
(8) in Knowle.
day ward number
2020-12-05 1 5
2020-12-10 1 5
2020-12-15 1 5
Seeding using three isoformat dates dates (5th, 10th and 15th December 2020), all in ward 1, seeding 5 infections each day.