robertness

Global Flag,Major Flag,SMC Flag,Commercial Flag,IT Spend,Employee Count,PC Count,Size,Tech Support,Discount,New Engagement Strategy,New Product Adoption,Planning Summit,Revenue,Direct Treatment Effect: Tech Support,Total Treatment Effect: Tech Support,Direct Treatment Effect: Discount,Total Treatment Effect: Discount,Direct Treatment Effect: New Engagement Strategy,Total Treatment Effect: New Engagement Strategy
0,0,1,1,72490,27,25,220417,1,0,0,1,0,39691.32,9408.34,10332.57,11020.85,11020.85,0,0
0,0,1,0,5985,19,12,12510,1,0,1,1,1,14492.54,5250.2,6174.43,625.5,625.5,0,0
0,0,1,0,34399,82,55,123530,0,0,0,0,0,16215.76,7470.6,8394.83,6176.5,6176.5,0,0
0,1,1,1,29705,103,58,79297,0,1,1,1,1,25426.04,6585.9400000000005,7510.17,3964.8500000000004,3964.8500000000004,0,0
0,0,1,1,13498,60,26,25472,1,0,1,1,1,21005.51,5509.44,6433.67,1273.6000000000001,1273.6000000000001,0,0
0,0,1,0,1625,149,108,6435,0,1,1,1,0,8637.43,5128.7,6052.93,321.75,321.75,0,0
0,0,0,0,64679,52,25,188629,1,0,0,1,0,32006.67,8772.58,9696.81,9431.45,9431

robertness

<?xml version="1.0" encoding="UTF-8"?>
<sbml xmlns="http://www.sbml.org/sbml/level2" level="2" version="1">
  <model id="BSMod02" name="Bacillus subtilis motility with GFP">
    <listOfUnitDefinitions>
      <unitDefinition id="substance">
        <listOfUnits>
          <unit kind="item" exponent="1" scale="0" multiplier="1"/>
        </listOfUnits>
      </unitDefinition>
    </listOfUnitDefinitions>

robertness

# Test that: constraints used to build equivalence class are stored in object
bl = arcs(nets$long)[1:3, c(2, 1)]
colnames(bl) = c("from", "to")
wl = arcs(nets$long)[4:5, ] 
colnames(wl) = c("from", "to")
beta = as.data.frame(arcs(nets$long)[6:7, ])
beta$prob = runif(nrow(beta))
exp = c("A", "G")
x = ctsdag(nets$long, exp, beta = beta, whitelist = wl, blacklist = bl)
stopifnot(names(x$constraints) == c("exp", "beta"))

robertness

    def summary(self):
        """
        Return parameter-level summary of the fit results.

        Returns:
            (pd.DataFrame): A dataframe with parameter-level summaries of the posterior.  
            Note column 'n_eff' corresponds to effective sample size and 'Rhat' 
            is a convergence metric.  See Stan docs for details.

        """

robertness

#LSMeams-like model
\begin{align*} 
\epsilon_{aijs} &\overset{\text{iid}}{\sim} N(0, \sigma)\\
Y_{aijs} &= \mu_s + R_{i} + C_{j} + A_{a} + R_{i}C_{j} + R^2_{i} + C^2_{j} + A^2_{a} + C_{j}A_{a} + R_{i}A_{a} + \lambda \left \| \vec{\theta}\  \right \|_2+ \epsilon_{aijs}
\end{align}

# Trent's model
\begin{align*}
R_{i} &\overset{\text{iid}}{\sim} N(0, 1), \ C_{j} \overset{\text{iid}}{\sim} N(0, 1)\\
\mu_{p(s)} &\overset{\text{iid}}{\sim} N(\mu, \sigma) \\ 

robertness

test.error = function(object){
  error_result = tryCatch({
    object
    FALSE
  }, error = function(e){
    TRUE
  })
  stopifnot(error_result)
}
expect.error <- function(object, regexp = NULL) {

robertness

stan_str <- "
data {
  int N;
  int D;
  int Y[N];
  int M[N];
  vector[D] X[N];
}
parameters {
  matrix[1, D] beta;

robertness

stan_str <- "
data {
  int N;
  int D;
  int Y[N];
  int M[N];
  vector[D] X[N];
}
parameters {
  matrix[1, D] beta;

robertness

library(faraway)
library(bnlearn)
data(femsmoke)
mdl <- empty.graph(c("dead", "age", "smoker"))
modelstring(mdl) <- "[dead|age:smoker][age][smoker]"
counts <- xtabs(y ~ dead + age + smoker, femsmoke) 
cpt_age <- margin.table(counts, 2) %>% prop.table 
cpt_smoker <- margin.table(counts, 3) %>% prop.table 
cpt_dead <- counts
cpt_dead[, , 1] <- prop.table(counts[, , 1], 2) 

robertness

#ifdef SIZE_DEFINITIONS
#define N_METABS 21
#define N_ODE_METABS 0
#define N_INDEP_METABS 8
#define N_COMPARTMENTS 1
#define N_GLOBAL_PARAMS 40
#define N_KIN_PARAMS 0
#define N_REACTIONS 20

#define N_ARRAY_SIZE_P  41	// number of parameters