tests/testthat/test-ssa.R
In rcannood/GillespieSSA2: Gillespie's Stochastic Simulation Algorithm for Impatient People
context("ssa simulation")
params <- c(
"a" = 1,
"b" = 2,
"c" = 3,
"d" = 4.5321
)
state <- c(
"a_very_long_state_name_value_is_just_to_try_and_see_if_it_works" = 14,
"short_one" = 7,
"x" = 1,
"y" = 0,
"z" = 0
)
reactions <- list(
reaction("a", c(x = +1), "react1"),
reaction("b", c(x = +2), "react2"),
reaction("c", c(y = +3), "react3"),
reaction("d", c(z = +4), "react4"),
reaction("a_very_long_state_name_value_is_just_to_try_and_see_if_it_works", c(x = +5, short_one = +1), "react5"),
reaction("short_one", c(short_one = -1, z = 1), "react6"),
reaction("x", c(x = -1), "react7"),
reaction("y", c(y = -1), "react8"),
reaction("z", c(z = -1), "react9"),
reaction(
propensity = "a * b * a_very_long_state_name_value_is_just_to_try_and_see_if_it_works * short_one / c / d",
effect = c(x = 3, y = 1),
name = "react10"
),
reaction(
propensity = "buf1 = a + 1; buf2 = buf1 + 1; buf3 = buf2 + 1; buf4 = buf3 + 1; buf4 + 1",
effect = c(y = +1, z = +1),
name = "react11"
),
reaction(
propensity = ~ (a * b + c) / d,
effect = c(y = +1, z = +1),
name = "react12"
)
)
expected_prop <- with(as.list(c(params, state)), c(
params,
state,
a * b * a_very_long_state_name_value_is_just_to_try_and_see_if_it_works * short_one / c / d,
a + 5,
(a * b + c) / d
)) %>%
unname()
comp_reac <- compile_reactions(
reactions = reactions,
state_ids = names(state),
params = params,
hardcode_params = FALSE
)
test_that("perform manual steps", {
sim <- create_simulation(
compiled_reactions = comp_reac,
params = params,
method_ptr = ssa_exact()$factory(),
initial_state = state,
final_time = 1,
log_buffer = TRUE,
log_propensity = TRUE,
log_firings = TRUE,
census_interval = .01,
max_walltime = 10,
stop_on_neg_state = FALSE,
sim_name = "test",
verbose = TRUE,
console_interval = 1
)
sim$reset()
# check variables before simulation
expect_equal(sim$all_zero_propensity, FALSE)
expect_equal(sim$all_zero_state, FALSE)
expect_equal(sim$buffer, 2:5)
expect_equal(sim$census_interval, 0.01)
expect_equal(sim$console_interval, 1)
expect_equal(sim$dfirings, rep(0, length(reactions)))
expect_equal(sim$dstate, rep(0, length(state)))
expect_equal(sim$dtime, 0)
expect_equal(sim$dtime_mean, 0)
expect_equal(sim$dtime_sd, 0)
expect_equal(sim$final_time, 1)
expect_equal(sim$firings, rep(0, length(reactions)))
expect_equal(sim$firings_mean, 0)
expect_equal(sim$firings_sd, 0)
expect_equal(sim$initial_state, state)
expect_equal(sim$log_buffer, TRUE)
expect_equal(sim$log_firings, TRUE)
expect_equal(sim$log_propensity, TRUE)
expect_equal(sim$max_walltime, 10)
expect_equal(sim$negative_propensity, FALSE)
expect_equal(sim$negative_state, FALSE)
expect_equal(sim$nu_i, comp_reac$state_change@i)
expect_equal(sim$nu_p, comp_reac$state_change@p)
expect_equal(sim$nu_x, comp_reac$state_change@x)
expect_equal(sim$num_steps, 0)
expect_equal(nrow(sim$output_buffer), 10)
expect_equal(ncol(sim$output_buffer), 4)
expect_equal(sim$output_buffer[1,], 2:5)
expect_true(all(sim$output_buffer[-1,] == 0))
expect_true(all(sim$output_firings == 0))
expect_equal(sim$output_nexti, 1)
expect_equal(nrow(sim$output_propensity), 10)
expect_equal(ncol(sim$output_propensity), length(reactions))
expect_equal(sim$output_propensity[1, ], expected_prop)
expect_true(all(sim$output_propensity[-1, ] == 0))
expect_equal(nrow(sim$output_state), 10)
expect_equal(ncol(sim$output_state), length(state))
expect_equal(sim$output_state[1, ], state %>% unname)
expect_true(all(sim$output_state[-1, ] == 0))
expect_true(all(sim$output_time == 0))
expect_equal(sim$params, params)
expect_equal(sim$propensity, expected_prop)
expect_equal(sim$sim_name, "test")
expect_equal(sim$sim_time, 0)
expect_equal(sim$state, state %>% unname)
expect_equal(sim$stop_on_neg_state, FALSE)
expect_equal(sim$verbose, TRUE)
sim$make_step()
expect_length(sim$dfirings, length(reactions))
expect_equal(sum(sim$dfirings == 1), 1L)
expect_equal(sum(sim$dfirings == 0), length(reactions) - 1)
expect_length(sim$dstate, length(state))
expect_is(sim$dstate, "numeric")
ix <- which(sim$dfirings == 1)
expect_equal(sim$dstate, comp_reac$state_change[, ix])
expect_length(sim$dtime, 1)
expect_is(sim$dtime, "numeric")
expect_gt(sim$dtime, 0)
expect_lte(sim$dtime, sum(expected_prop))
expect_equal(sim$dtime_mean, sim$dtime)
expect_equal(sim$firings_mean, 1)
sim$calculate_propensity()
# try to find out whether code might cause segfaults to occur
while (
sim$num_steps < 1000 &&
!sim$negative_propensity &&
!sim$negative_state &&
!sim$all_zero_propensity &&
!sim$all_zero_state
) {
sim$make_step()
sim$calculate_propensity()
}
expect_false(sim$negative_propensity)
expect_false(sim$negative_state)
expect_false(sim$all_zero_propensity)
expect_false(sim$all_zero_state)
})
test_that("perform simulation with Rcpp function", {
sim <- create_simulation(
compiled_reactions = comp_reac,
params = params,
method_ptr = ssa_exact()$factory(),
initial_state = state,
final_time = 1,
log_buffer = TRUE,
log_propensity = TRUE,
log_firings = TRUE,
census_interval = .01,
max_walltime = 10,
stop_on_neg_state = FALSE,
sim_name = "test",
verbose = TRUE,
console_interval = 1
)
sim$reset()
# let simulator run freely
expect_output(
{
sim$run()
},
regexp = "Running SSA exact"
)
# check simulator values again
expect_equal(sim$all_zero_propensity, FALSE)
expect_equal(sim$all_zero_state, FALSE)
expect_equal(sim$buffer, 2:5)
expect_equal(sim$census_interval, 0.01)
expect_equal(sim$console_interval, 1)
expect_equal(sum(sim$dfirings), 1)
expect_equal(sum(sim$dfirings != 0), 1)
ix <- which(sim$dfirings != 0)
exp_dstate <- comp_reac$state_change[,ix]
expect_equal(sim$dstate, exp_dstate)
expect_gt(sim$dtime, 0)
expect_gt(sim$dtime_mean, 0)
expect_gt(sim$dtime_sd, 0)
expect_gt(sim$final_time, .99)
expect_equal(sim$firings, rep(0, length(reactions)))
expect_equal(sim$firings_mean, 1)
expect_equal(sim$firings_sd, 0)
expect_equal(sim$initial_state, state)
expect_equal(sim$log_buffer, TRUE)
expect_equal(sim$log_firings, TRUE)
expect_equal(sim$log_propensity, TRUE)
expect_equal(sim$max_walltime, 10)
expect_equal(sim$negative_propensity, FALSE)
expect_equal(sim$negative_state, FALSE)
expect_equal(sim$nu_i, comp_reac$state_change@i)
expect_equal(sim$nu_p, comp_reac$state_change@p)
expect_equal(sim$nu_x, comp_reac$state_change@x)
expect_gte(sim$num_steps, 0)
expect_gte(sim$output_nexti, 1)
expect_equal(nrow(sim$output_buffer), sim$output_nexti)
expect_equal(ncol(sim$output_buffer), 4)
expect_true(all(apply(sim$output_buffer, 1, function(x) all(x == 2:5))))
expect_equal(nrow(sim$output_firings), sim$output_nexti)
expect_equal(ncol(sim$output_firings), length(reactions))
expect_true(all(rowSums(sim$output_firings[-1,]) >= 1))
expect_equal(nrow(sim$output_propensity), sim$output_nexti)
expect_equal(ncol(sim$output_propensity), length(reactions))
expect_equal(sim$output_propensity[1, ], expected_prop)
expect_equal(nrow(sim$output_state), sim$output_nexti)
expect_equal(ncol(sim$output_state), length(state))
expect_equal(sim$output_state[1, ], state %>% unname)
expect_equal(length(sim$output_time), sim$output_nexti)
expect_true(all(diff(sim$output_time) > 0))
expect_equal(sim$params, params)
exp_state <- state + (comp_reac$state_change %*% colSums(sim$output_firings))[,1]
exp_prop <- with(as.list(c(params, exp_state)), c(
params,
exp_state,
a * b * a_very_long_state_name_value_is_just_to_try_and_see_if_it_works * short_one / c / d,
a + 5,
(a * b + c) / d
)) %>%
unname()
expect_equal(sim$propensity, exp_prop)
expect_equal(sim$sim_name, "test")
expect_gte(sim$sim_time, 1)
expect_equal(sim$state, exp_state %>% unname)
expect_equal(sim$stop_on_neg_state, FALSE)
expect_equal(sim$verbose, TRUE)
stats <- sim$get_statistics()
expect_is(stats, "data.frame")
})
test_that("perform simulation with R function", {
expect_output(
{
out <- ssa(
initial_state = state,
reactions = reactions,
final_time = 1,
params = params,
method = ssa_exact(),
census_interval = .01,
max_walltime = 10,
stop_on_neg_state = TRUE,
log_propensity = TRUE,
log_firings = TRUE,
log_buffer = TRUE,
verbose = TRUE,
console_interval = 1,
sim_name = "test"
)
},
regexp = "Running SSA exact"
)
# check simulator values again
expect_equal(ncol(out$buffer), 4)
expect_true(all(apply(out$buffer, 1, function(x) all(x == 2:5))))
expect_equal(nrow(out$firings), nrow(out$buffer))
expect_equal(ncol(out$firings), length(reactions))
expect_true(all(rowSums(out$firings[-1,]) >= 1))
expect_equal(nrow(out$propensity), nrow(out$buffer))
expect_equal(ncol(out$propensity), length(reactions))
expect_equal(out$propensity[1, ], expected_prop %>% setNames(comp_reac$reaction_ids))
expect_equal(nrow(out$state), nrow(out$buffer))
expect_equal(ncol(out$state), length(state))
expect_equal(out$state[1, ], state)
expect_equal(length(out$time), nrow(out$buffer))
expect_true(all(diff(out$time) > 0))
exp_state <- state + (comp_reac$state_change %*% colSums(out$firings))[,1]
exp_prop <- with(as.list(c(params, exp_state)), c(
params,
exp_state,
a * b * a_very_long_state_name_value_is_just_to_try_and_see_if_it_works * short_one / c / d,
a + 5,
(a * b + c) / d
)) %>% setNames(comp_reac$reaction_ids)
expect_equal(out$propensity[nrow(out$propensity),], exp_prop)
expect_gte(out$time[[nrow(out$buffer)]], 1)
expect_equal(out$state[nrow(out$state), ], exp_state)
stats <- out$stats
expect_is(stats, "data.frame")
g <- plot_ssa(
out,
state = TRUE,
propensity = TRUE,
buffer = TRUE
)
expect_is(g, "ggplot")
g <- plot_ssa(
out,
state = FALSE,
propensity = FALSE,
buffer = TRUE,
geom = "step"
)
expect_is(g, "ggplot")
})
test_that("multiple runs produces different results", {
out1 <- ssa(
initial_state = state,
reactions = comp_reac,
final_time = 1,
params = params,
method = ssa_exact(),
census_interval = .01,
max_walltime = 10,
stop_on_neg_state = TRUE,
log_propensity = TRUE,
log_firings = TRUE,
log_buffer = TRUE,
verbose = TRUE,
console_interval = 1,
sim_name = "test"
)
out2 <- ssa(
initial_state = state,
reactions = comp_reac,
final_time = 1,
params = params,
method = ssa_exact(),
census_interval = .01,
max_walltime = 10,
stop_on_neg_state = TRUE,
log_propensity = TRUE,
log_firings = TRUE,
log_buffer = TRUE,
verbose = TRUE,
console_interval = 1,
sim_name = "test"
)
expect_false(
identical(out1$state, out2$state)
)
})
test_that("simulation when a negative propensity is generated", {
params <- c("a" = -1)
state <- c("x" = 1)
reactions <- list(
reaction("a", c(x = +1), "react1")
)
comp_reac <- compile_reactions(
reactions = reactions,
state_ids = names(state),
params = params
)
sim <- create_simulation(
compiled_reactions = comp_reac,
params = params,
initial_state = state,
final_time = 1
)
sim$state <- state
sim$calculate_propensity()
expect_equal(sim$propensity, -1)
expect_true(sim$negative_propensity)
})
test_that("simulation when a negative state value is generated", {
params <- c("a" = 1)
state <- c("x" = 1)
reactions <- list(
reaction("a", c(x = -1000), "react1")
)
comp_reac <- compile_reactions(
reactions = reactions,
state_ids = names(state),
params = params
)
sim <- create_simulation(
compiled_reactions = comp_reac,
params = params,
initial_state = state,
final_time = 1
)
sim$state <- state
sim$calculate_propensity()
sim$make_step()
expect_equal(sim$state %>% unname, -999)
expect_true(sim$negative_state)
})
rcannood/GillespieSSA2 documentation built on Jan. 26, 2023, 8:33 p.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
context("ssa simulation")
params <- c(
"a" = 1,
"b" = 2,
"c" = 3,
"d" = 4.5321
)
state <- c(
"a_very_long_state_name_value_is_just_to_try_and_see_if_it_works" = 14,
"short_one" = 7,
"x" = 1,
"y" = 0,
"z" = 0
)
reactions <- list(
reaction("a", c(x = +1), "react1"),
reaction("b", c(x = +2), "react2"),
reaction("c", c(y = +3), "react3"),
reaction("d", c(z = +4), "react4"),
reaction("a_very_long_state_name_value_is_just_to_try_and_see_if_it_works", c(x = +5, short_one = +1), "react5"),
reaction("short_one", c(short_one = -1, z = 1), "react6"),
reaction("x", c(x = -1), "react7"),
reaction("y", c(y = -1), "react8"),
reaction("z", c(z = -1), "react9"),
reaction(
propensity = "a * b * a_very_long_state_name_value_is_just_to_try_and_see_if_it_works * short_one / c / d",
effect = c(x = 3, y = 1),
name = "react10"
),
reaction(
propensity = "buf1 = a + 1; buf2 = buf1 + 1; buf3 = buf2 + 1; buf4 = buf3 + 1; buf4 + 1",
effect = c(y = +1, z = +1),
name = "react11"
),
reaction(
propensity = ~ (a * b + c) / d,
effect = c(y = +1, z = +1),
name = "react12"
)
)
expected_prop <- with(as.list(c(params, state)), c(
params,
state,
a * b * a_very_long_state_name_value_is_just_to_try_and_see_if_it_works * short_one / c / d,
a + 5,
(a * b + c) / d
)) %>%
unname()
comp_reac <- compile_reactions(
reactions = reactions,
state_ids = names(state),
params = params,
hardcode_params = FALSE
)
test_that("perform manual steps", {
sim <- create_simulation(
compiled_reactions = comp_reac,
params = params,
method_ptr = ssa_exact()$factory(),
initial_state = state,
final_time = 1,
log_buffer = TRUE,
log_propensity = TRUE,
log_firings = TRUE,
census_interval = .01,
max_walltime = 10,
stop_on_neg_state = FALSE,
sim_name = "test",
verbose = TRUE,
console_interval = 1
)
sim$reset()
# check variables before simulation
expect_equal(sim$all_zero_propensity, FALSE)
expect_equal(sim$all_zero_state, FALSE)
expect_equal(sim$buffer, 2:5)
expect_equal(sim$census_interval, 0.01)
expect_equal(sim$console_interval, 1)
expect_equal(sim$dfirings, rep(0, length(reactions)))
expect_equal(sim$dstate, rep(0, length(state)))
expect_equal(sim$dtime, 0)
expect_equal(sim$dtime_mean, 0)
expect_equal(sim$dtime_sd, 0)
expect_equal(sim$final_time, 1)
expect_equal(sim$firings, rep(0, length(reactions)))
expect_equal(sim$firings_mean, 0)
expect_equal(sim$firings_sd, 0)
expect_equal(sim$initial_state, state)
expect_equal(sim$log_buffer, TRUE)
expect_equal(sim$log_firings, TRUE)
expect_equal(sim$log_propensity, TRUE)
expect_equal(sim$max_walltime, 10)
expect_equal(sim$negative_propensity, FALSE)
expect_equal(sim$negative_state, FALSE)
expect_equal(sim$nu_i, comp_reac$state_change@i)
expect_equal(sim$nu_p, comp_reac$state_change@p)
expect_equal(sim$nu_x, comp_reac$state_change@x)
expect_equal(sim$num_steps, 0)
expect_equal(nrow(sim$output_buffer), 10)
expect_equal(ncol(sim$output_buffer), 4)
expect_equal(sim$output_buffer[1,], 2:5)
expect_true(all(sim$output_buffer[-1,] == 0))
expect_true(all(sim$output_firings == 0))
expect_equal(sim$output_nexti, 1)
expect_equal(nrow(sim$output_propensity), 10)
expect_equal(ncol(sim$output_propensity), length(reactions))
expect_equal(sim$output_propensity[1, ], expected_prop)
expect_true(all(sim$output_propensity[-1, ] == 0))
expect_equal(nrow(sim$output_state), 10)
expect_equal(ncol(sim$output_state), length(state))
expect_equal(sim$output_state[1, ], state %>% unname)
expect_true(all(sim$output_state[-1, ] == 0))
expect_true(all(sim$output_time == 0))
expect_equal(sim$params, params)
expect_equal(sim$propensity, expected_prop)
expect_equal(sim$sim_name, "test")
expect_equal(sim$sim_time, 0)
expect_equal(sim$state, state %>% unname)
expect_equal(sim$stop_on_neg_state, FALSE)
expect_equal(sim$verbose, TRUE)
sim$make_step()
expect_length(sim$dfirings, length(reactions))
expect_equal(sum(sim$dfirings == 1), 1L)
expect_equal(sum(sim$dfirings == 0), length(reactions) - 1)
expect_length(sim$dstate, length(state))
expect_is(sim$dstate, "numeric")
ix <- which(sim$dfirings == 1)
expect_equal(sim$dstate, comp_reac$state_change[, ix])
expect_length(sim$dtime, 1)
expect_is(sim$dtime, "numeric")
expect_gt(sim$dtime, 0)
expect_lte(sim$dtime, sum(expected_prop))
expect_equal(sim$dtime_mean, sim$dtime)
expect_equal(sim$firings_mean, 1)
sim$calculate_propensity()
# try to find out whether code might cause segfaults to occur
while (
sim$num_steps < 1000 &&
!sim$negative_propensity &&
!sim$negative_state &&
!sim$all_zero_propensity &&
!sim$all_zero_state
) {
sim$make_step()
sim$calculate_propensity()
}
expect_false(sim$negative_propensity)
expect_false(sim$negative_state)
expect_false(sim$all_zero_propensity)
expect_false(sim$all_zero_state)
})
test_that("perform simulation with Rcpp function", {
sim <- create_simulation(
compiled_reactions = comp_reac,
params = params,
method_ptr = ssa_exact()$factory(),
initial_state = state,
final_time = 1,
log_buffer = TRUE,
log_propensity = TRUE,
log_firings = TRUE,
census_interval = .01,
max_walltime = 10,
stop_on_neg_state = FALSE,
sim_name = "test",
verbose = TRUE,
console_interval = 1
)
sim$reset()
# let simulator run freely
expect_output(
{
sim$run()
},
regexp = "Running SSA exact"
)
# check simulator values again
expect_equal(sim$all_zero_propensity, FALSE)
expect_equal(sim$all_zero_state, FALSE)
expect_equal(sim$buffer, 2:5)
expect_equal(sim$census_interval, 0.01)
expect_equal(sim$console_interval, 1)
expect_equal(sum(sim$dfirings), 1)
expect_equal(sum(sim$dfirings != 0), 1)
ix <- which(sim$dfirings != 0)
exp_dstate <- comp_reac$state_change[,ix]
expect_equal(sim$dstate, exp_dstate)
expect_gt(sim$dtime, 0)
expect_gt(sim$dtime_mean, 0)
expect_gt(sim$dtime_sd, 0)
expect_gt(sim$final_time, .99)
expect_equal(sim$firings, rep(0, length(reactions)))
expect_equal(sim$firings_mean, 1)
expect_equal(sim$firings_sd, 0)
expect_equal(sim$initial_state, state)
expect_equal(sim$log_buffer, TRUE)
expect_equal(sim$log_firings, TRUE)
expect_equal(sim$log_propensity, TRUE)
expect_equal(sim$max_walltime, 10)
expect_equal(sim$negative_propensity, FALSE)
expect_equal(sim$negative_state, FALSE)
expect_equal(sim$nu_i, comp_reac$state_change@i)
expect_equal(sim$nu_p, comp_reac$state_change@p)
expect_equal(sim$nu_x, comp_reac$state_change@x)
expect_gte(sim$num_steps, 0)
expect_gte(sim$output_nexti, 1)
expect_equal(nrow(sim$output_buffer), sim$output_nexti)
expect_equal(ncol(sim$output_buffer), 4)
expect_true(all(apply(sim$output_buffer, 1, function(x) all(x == 2:5))))
expect_equal(nrow(sim$output_firings), sim$output_nexti)
expect_equal(ncol(sim$output_firings), length(reactions))
expect_true(all(rowSums(sim$output_firings[-1,]) >= 1))
expect_equal(nrow(sim$output_propensity), sim$output_nexti)
expect_equal(ncol(sim$output_propensity), length(reactions))
expect_equal(sim$output_propensity[1, ], expected_prop)
expect_equal(nrow(sim$output_state), sim$output_nexti)
expect_equal(ncol(sim$output_state), length(state))
expect_equal(sim$output_state[1, ], state %>% unname)
expect_equal(length(sim$output_time), sim$output_nexti)
expect_true(all(diff(sim$output_time) > 0))
expect_equal(sim$params, params)
exp_state <- state + (comp_reac$state_change %*% colSums(sim$output_firings))[,1]
exp_prop <- with(as.list(c(params, exp_state)), c(
params,
exp_state,
a * b * a_very_long_state_name_value_is_just_to_try_and_see_if_it_works * short_one / c / d,
a + 5,
(a * b + c) / d
)) %>%
unname()
expect_equal(sim$propensity, exp_prop)
expect_equal(sim$sim_name, "test")
expect_gte(sim$sim_time, 1)
expect_equal(sim$state, exp_state %>% unname)
expect_equal(sim$stop_on_neg_state, FALSE)
expect_equal(sim$verbose, TRUE)
stats <- sim$get_statistics()
expect_is(stats, "data.frame")
})
test_that("perform simulation with R function", {
expect_output(
{
out <- ssa(
initial_state = state,
reactions = reactions,
final_time = 1,
params = params,
method = ssa_exact(),
census_interval = .01,
max_walltime = 10,
stop_on_neg_state = TRUE,
log_propensity = TRUE,
log_firings = TRUE,
log_buffer = TRUE,
verbose = TRUE,
console_interval = 1,
sim_name = "test"
)
},
regexp = "Running SSA exact"
)
# check simulator values again
expect_equal(ncol(out$buffer), 4)
expect_true(all(apply(out$buffer, 1, function(x) all(x == 2:5))))
expect_equal(nrow(out$firings), nrow(out$buffer))
expect_equal(ncol(out$firings), length(reactions))
expect_true(all(rowSums(out$firings[-1,]) >= 1))
expect_equal(nrow(out$propensity), nrow(out$buffer))
expect_equal(ncol(out$propensity), length(reactions))
expect_equal(out$propensity[1, ], expected_prop %>% setNames(comp_reac$reaction_ids))
expect_equal(nrow(out$state), nrow(out$buffer))
expect_equal(ncol(out$state), length(state))
expect_equal(out$state[1, ], state)
expect_equal(length(out$time), nrow(out$buffer))
expect_true(all(diff(out$time) > 0))
exp_state <- state + (comp_reac$state_change %*% colSums(out$firings))[,1]
exp_prop <- with(as.list(c(params, exp_state)), c(
params,
exp_state,
a * b * a_very_long_state_name_value_is_just_to_try_and_see_if_it_works * short_one / c / d,
a + 5,
(a * b + c) / d
)) %>% setNames(comp_reac$reaction_ids)
expect_equal(out$propensity[nrow(out$propensity),], exp_prop)
expect_gte(out$time[[nrow(out$buffer)]], 1)
expect_equal(out$state[nrow(out$state), ], exp_state)
stats <- out$stats
expect_is(stats, "data.frame")
g <- plot_ssa(
out,
state = TRUE,
propensity = TRUE,
buffer = TRUE
)
expect_is(g, "ggplot")
g <- plot_ssa(
out,
state = FALSE,
propensity = FALSE,
buffer = TRUE,
geom = "step"
)
expect_is(g, "ggplot")
})
test_that("multiple runs produces different results", {
out1 <- ssa(
initial_state = state,
reactions = comp_reac,
final_time = 1,
params = params,
method = ssa_exact(),
census_interval = .01,
max_walltime = 10,
stop_on_neg_state = TRUE,
log_propensity = TRUE,
log_firings = TRUE,
log_buffer = TRUE,
verbose = TRUE,
console_interval = 1,
sim_name = "test"
)
out2 <- ssa(
initial_state = state,
reactions = comp_reac,
final_time = 1,
params = params,
method = ssa_exact(),
census_interval = .01,
max_walltime = 10,
stop_on_neg_state = TRUE,
log_propensity = TRUE,
log_firings = TRUE,
log_buffer = TRUE,
verbose = TRUE,
console_interval = 1,
sim_name = "test"
)
expect_false(
identical(out1$state, out2$state)
)
})
test_that("simulation when a negative propensity is generated", {
params <- c("a" = -1)
state <- c("x" = 1)
reactions <- list(
reaction("a", c(x = +1), "react1")
)
comp_reac <- compile_reactions(
reactions = reactions,
state_ids = names(state),
params = params
)
sim <- create_simulation(
compiled_reactions = comp_reac,
params = params,
initial_state = state,
final_time = 1
)
sim$state <- state
sim$calculate_propensity()
expect_equal(sim$propensity, -1)
expect_true(sim$negative_propensity)
})
test_that("simulation when a negative state value is generated", {
params <- c("a" = 1)
state <- c("x" = 1)
reactions <- list(
reaction("a", c(x = -1000), "react1")
)
comp_reac <- compile_reactions(
reactions = reactions,
state_ids = names(state),
params = params
)
sim <- create_simulation(
compiled_reactions = comp_reac,
params = params,
initial_state = state,
final_time = 1
)
sim$state <- state
sim$calculate_propensity()
sim$make_step()
expect_equal(sim$state %>% unname, -999)
expect_true(sim$negative_state)
})
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.