|
cluster.order()
|
order cluster by increasing wage |
|
dstats()
|
provides statistics |
|
em.control()
|
Create a control structure for running EM algorithms |
|
estimation.threeSided.model()
|
Mixture model estimation of three sided model |
|
get.largest.conset.fid()
|
some work on trace formula
Extracts the largest connected set from data using f1,f2
movers |
|
get.largest.leaveoutset.fid()
|
Extracts the largest leave-out connected set from data using f1,f2
movers |
|
get.sample.stats()
|
computes simple statistics |
|
grouping.append()
|
Append result of a grouping to a data-set |
|
grouping.classify.once()
|
clusters firms based on their cross-sectional wage distributions |
|
grouping.classify()
|
clusters firms based on their cross-sectional wage distributions |
|
grouping.computeobj()
|
Compute the objective function of the clustering |
|
grouping.getMeasures.em()
|
Get the three sided measures for inputs in clustering |
|
grouping.getMeasures()
|
Extract the measurement matrix to be given to the
classification algorithm |
|
grouping.infos()
|
extract information |
|
grouping.makefiner()
|
Gives their won cluster to firm with many movers |
|
jdata.prepare()
|
Prepare the data for BLM from an employer-employee matched data |
|
kmeansW.repeat()
|
internal function that runs Kmean with multiple starting values |
|
lin.proja()
|
Generate a linear projection decomposition for the model
with continuous worker hetergoneity |
|
lin.projax()
|
Computes the linear projection using X |
|
lin.projx()
|
Computes the linear projection using X |
|
lognormpdf()
|
functions for em |
|
logRowSumExp()
|
logsumexp function by Row |
|
logsumexp()
|
logsumexp function |
|
m2.firmfe.pen()
|
Ridge AKM |
|
m2.get.pk_unc()
|
Returns the uconditional type probability in the crossection |
|
m2.mixt.estimate.all()
|
Estimates the static mixture model on 2 periods |
|
m2.mixt.meaneffect()
|
Compute mean effects |
|
m2.mixt.movers()
|
Estimates the static model parameters for movers |
|
m2.mixt.new()
|
create a random model for EM with three sided
endogenous mobility with multinomial pr |
|
m2.mixt.pplot()
|
plots the proportions of a model |
|
m2.mixt.simulate.movers()
|
Using the model, simulates a dataset of movers |
|
m2.mixt.simulate.sim.clust()
|
Simulates data (movers and stayers) |
|
m2.mixt.simulate.sim()
|
Simulates data (movers and stayers) and attached firms ids. Firms have all same expected size. |
|
m2.mixt.simulate.stayers()
|
Using the model, simulates a dataset of stayers. |
|
m2.mixt.simulate.stayers.withx()
|
Using the model, simulates a dataset of stayers. |
|
m2.mixt.stayers()
|
use the marginal distributions to extract type distributions
within each cluster and observable characteristics |
|
m2.mixt.transform.data()
|
Data tranformation ( 3d array to 2d array) |
|
m2.mixt.transform.model()
|
Model transformation for solver (3d array object to 2d array objects)
used in multicore effitiency |
|
m2.mixt.vdec()
|
Computes the variance decomposition by simulation |
|
m2.mixt.wplot()
|
plots the wages of a model |
|
m2.movers.checkfit()
|
check the fit in the movers/stayers using imputed data |
|
m2.stayers.checkfit()
|
check the fit in the movers/stayers using imputed data |
|
m2.trace.estimate()
|
gets the connected set, then |
|
m2.trace.new()
|
create a model for testing trace estimation |
|
m2.trace.simulate.old()
|
simulates for trace estimation |
|
m2.trace.simulate()
|
simulates for trace estimation |
|
mcast()
|
creates a matrix and fill it using a data.table
in contrast to acast, it creates all rows and cols (even if there is not data) |
|
model(<connectiveness>)
|
Computes graph connectedness among the movers
within each type and returns the smalless value |
|
ModelInitializer()
|
Prepare the finction for test run
intialize the model parameters (helper function) |
|
plot(<trquant>)
|
plots the conditional quantile distribution for each transitions (l,l') |
|
plot(<vaeffect>)
|
we want to look at the effect of
movers on value added. |
|
plot(<wage>)
|
plot the mean wage at origin conditional on where it is coming from |
|
sample.stats()
|
compute some stats on data |
|
sColSums()
|
Sparse colSums |
|
set.solver.controls()
|
Set the solver controls |
|
Simulate.data.threeSided()
|
Simulate the data consiting of three sided heterogeniety |
|
spread()
|
this is a utility function to generate
multidimensional arrays - like the spread function in fortran |
|
sRowSums()
|
Sparse rowSums |
|
threeSided.Clustering()
|
Clustering (Step 1 of the estimation: Seep paper for details) |
|
threeSided.means.plot()
|
Estimated means plot |
|
threeSided.proportion.plot()
|
Proportion plot |
|
vcast()
|
creates a vector and fill it using a data.table
in contrast to acast, it creates all elements |
|
wt.cov()
|
Weighted covariance |