FIESTA - Model-Assisted Estimators

Model-Assisted (MA) module overview

FIESTA’s Model-Assisted (MA) module calculates population estimates and their sampling errors by taking advantage of available model-assisted survey estimators from the mase R package (McConville, et al. 2018). These estimators can use a variety of auxiliary data to build models and predict over a response variable of interest, while using a bias-correction term so that the bias of the model does not depend on model mis-specification.

Functions in FIESTA used for fitting model-assisted estimators include the modMAarea function for area estimates and modMAtree for tree estimates. The modMApop function is used to get population data needed for model-assisted estimation. Below is a description and table of contents for the sections related to these functions:

FUNCTION DESCRIPTION
modMApop Creates population data for model-assisted estimation.
modMAarea Produces area level estimates through model-assisted estimation.
modMAtree Produces tree level estimates through model-assisted estimation.

Objective of tutorial

The main objective of this tutorial is to demonstrate how to use FIESTA for generating estimates using estimators from mase. The model-assisted estimators can be used with FIA’s standard state-level population data (i.e, Evaluation) from the FIA database (FIADB) and also population data from a custom boundary.

The following examples are for generating estimates and estimated variances using standard FIA Evaluation data from FIA’s National database, with custom Estimation unit and Stratification information. The examples use data from three inventory years of field measurements in the state of Wyoming, from FIADB_1.7.2.00, last updated June 20, 2018, downloaded on June 25, 2018 and stored as internal data objects in FIESTA.

Example data - Wyoming (WY), Inventory Years 2011-2012

View MA Example Data
Data Frame Description
WYplt WY plot-level data
WYcond WY condition-level data
WYtree WY tree-level data
External data Description
WYbighorn_adminbnd.shp Polygon shapefile of WY Bighorn National Forest Administrative boundary*
WYbighorn_districtbnd.shp Polygon shapefile of WY Bighorn National Forest District boundaries**
WYbighorn_forest_nonforest_250m.tif GeoTIFF raster of predicted forest/nonforest (1/0) for stratification***
WYbighorn_dem_250m.img Erdas Imagine raster of elevation change, in meters****

*USDA Forest Service, Automated Lands Program (ALP). 2018. S_USA.AdministrativeForest (http://data.fs.usda.gov/geodata/edw). Description: An area encompassing all the National Forest System lands administered by an administrative unit. The area encompasses private lands, other governmental agency lands, and may contain National Forest System lands within the proclaimed boundaries of another administrative unit. All National Forest System lands fall within one and only one Administrative Forest Area.

**USDA Forest Service, Automated Lands Program (ALP). 2018. S_USA.RangerDistrict (http://data.fs.usda.gov/geodata/edw). Description: A depiction of the boundary that encompasses a Ranger District.

***Based on MODIS-based classified map resampled from 250m to 500m resolution and reclassified from 3 to 2 classes: 1:forest; 2:nonforest. Projected in Albers Conical Equal Area, Datum NAD27 (Ruefenacht et al. 2008). Clipped to extent of WYbighorn_adminbnd.shp.

****USGS National Elevation Dataset (NED), resampled from 30m resolution to 250m. Projected in Albers Conical Equal Area, Datum NAD27 (U.S. Geological Survey 2017). Clipped to boundary of WYbighorn_adminbnd.shp.

Set up

First, you’ll need to load the FIESTA library:

library(FIESTA)

Next, you’ll need to set up an “outfolder”. This is just a file path to a folder where you’d like FIESTA to send your data output. For our purposes in this vignette, we have saved our outfolder file path as the outfolder object in a temporary directory. We also set a few default options preferred for this vignette.

outfolder <- tempdir()

Get data for examples

View Getting Data

Now that we’ve loaded FIESTA and setup our outfolder, we can retrieve the data needed to run the examples. First, we point to some external data and predictor layers stored in FIESTA and derive new predictor layers using the terra package.

# File names for external spatial data
WYbhfn <- system.file("extdata", "sp_data/WYbighorn_adminbnd.shp",
                      package = "FIESTA")
WYbhdistfn <- system.file("extdata", "sp_data/WYbighorn_districtbnd.shp",
                          package = "FIESTA")

## predictor variables
fornffn <- system.file("extdata", "sp_data/WYbighorn_forest_nonforest_250m.tif",
                       package = "FIESTA")
demfn <- system.file("extdata", "sp_data/WYbighorn_dem_250m.img",
                     package = "FIESTA")

# Derive new predictor layers from dem
library(terra)
dem <- rast(demfn)
slpfn <- paste0(outfolder, "/WYbh_slp.img")
slp <- terra::terrain(dem,
                      v = "slope",
                      unit = "degrees",
                      filename = slpfn, 
                      overwrite = TRUE,
                      NAflag = -99999.0)
aspfn <- paste0(outfolder, "/WYbh_asp.img")
asp <- terra::terrain(dem,
                      v = "aspect",
                      unit = "degrees", 
                      filename = aspfn,
                      overwrite = TRUE,
                      NAflag = -99999.0)

Next, we can get our FIA plot data and set up our auxiliary data. We can get our FIA plot data with the spMakeSpatialPoints function from FIESTA. For more information on how to use this function, please see the sp vignette included with FIESTA (link).

WYspplt <- spMakeSpatialPoints(
  xyplt = WYplt,
  xy.uniqueid = "CN",
  xvar = "LON_PUBLIC",
  yvar = "LAT_PUBLIC",
  xy.crs = 4269
)

rastlst.cont <- c(demfn, slpfn, aspfn)
rastlst.cont.name <- c("dem", "slp", "asp")
rastlst.cat <- fornffn
rastlst.cat.name <- "fornf"

Next, we must generate dataset for model-assisted estimation. We can do this with the spGetAuxiliary function from FIESTA. Again, see the sp vignette for further information on this function.

modeldat <- spGetAuxiliary(
  xyplt = WYspplt,
  uniqueid = "CN",
  unit_layer = WYbhfn,
  unitvar = NULL,
  rastlst.cont = rastlst.cont,
  rastlst.cont.name = rastlst.cont.name,
  rastlst.cat = rastlst.cat,
  rastlst.cat.name = rastlst.cat.name,
  rastlst.cont.stat = "mean",
  asptransform = TRUE,
  rast.asp = aspfn,
  keepNA = FALSE,
  showext = FALSE,
  savedata = FALSE)
str(modeldat, max.level = 1)
output
## List of 12
##  $ unitvar       : chr "ONEUNIT"
##  $ pltassgn      :'data.frame':  56 obs. of  25 variables:
##  $ pltassgnid    : chr "CN"
##  $ unitarea      :'data.frame':  1 obs. of  2 variables:
##  $ areavar       : chr "ACRES_GIS"
##  $ unitzonal     :'data.frame':  1 obs. of  8 variables:
##  $ inputdf       :Classes 'data.table' and 'data.frame': 4 obs. of  7 variables:
##   ..- attr(*, ".internal.selfref")=<externalptr> 
##  $ prednames     : chr [1:5] "dem" "slp" "asp_cos" "asp_sin" ...
##  $ zonalnames    : chr [1:7] "dem" "slp" "asp_cos" "asp_sin" ...
##  $ predfac       : chr "fornf"
##  $ npixelvar     : chr "npixels"
##  $ predfac.levels:List of 1

Examples

modMApop

Example 1: Creating our population dataset with modMApop

View Example

We can create our population data for model-assisted estimation. To do so, we use the modMApop function in FIESTA. We must assign our population tables with the popTabs argument (and unique identifiers for these tables with the popTabIDs argument if they are not the default), the plot assignment with the pltassgn argument, and in auxiliary dataset we just created with the auxdat argument. The spGetAuxiliary function has done much of the hard work for us so far, so we can just run a simple call to modMApop:

MApopdat <- modMApop(popTabs = list(tree = WYtree, cond = WYcond),
                     auxdat = modeldat)

Note that the modMApop function returns a list with lots of information and data for us to use. For a quick look at what this list includes we can use the str function:

str(MApopdat, max.level = 1)
output
## List of 26
##  $ module     : chr "MA"
##  $ popType    : chr "VOL"
##  $ condx      :Classes 'data.table' and 'data.frame':    66 obs. of  17 variables:
##   ..- attr(*, "sorted")= chr [1:2] "PLT_CN" "CONDID"
##   ..- attr(*, ".internal.selfref")=<externalptr> 
##  $ pltcondx   :Classes 'data.table' and 'data.frame':    66 obs. of  29 variables:
##   ..- attr(*, ".internal.selfref")=<externalptr> 
##   ..- attr(*, "sorted")= chr [1:2] "PLT_CN" "CONDID"
##  $ cuniqueid  : chr "PLT_CN"
##  $ condid     : chr "CONDID"
##  $ ACI.filter : chr "COND_STATUS_CD == 1"
##  $ unitarea   :Classes 'data.table' and 'data.frame':    1 obs. of  2 variables:
##   ..- attr(*, ".internal.selfref")=<externalptr> 
##   ..- attr(*, "sorted")= chr "ONEUNIT"
##  $ areavar    : chr "ACRES_GIS"
##  $ areaunits  : chr "acres"
##  $ unitvar    : chr "ONEUNIT"
##  $ unitvars   : chr "ONEUNIT"
##  $ unitlut    :Classes 'data.table' and 'data.frame':    1 obs. of  7 variables:
##   ..- attr(*, ".internal.selfref")=<externalptr> 
##   ..- attr(*, "sorted")= chr "ONEUNIT"
##  $ plotsampcnt:'data.frame': 2 obs. of  3 variables:
##  $ condsampcnt:'data.frame': 4 obs. of  3 variables:
##  $ npixels    :Classes 'data.table' and 'data.frame':    1 obs. of  2 variables:
##   ..- attr(*, ".internal.selfref")=<externalptr> 
##  $ npixelvar  : chr "npixels"
##  $ states     : chr "Wyoming"
##  $ invyrs     :List of 1
##  $ estvar.area: chr "CONDPROP_ADJ"
##  $ adj        : chr "plot"
##  $ treex      :Classes 'data.table' and 'data.frame':    1691 obs. of  21 variables:
##   ..- attr(*, ".internal.selfref")=<externalptr> 
##   ..- attr(*, "sorted")= chr [1:2] "PLT_CN" "CONDID"
##  $ tuniqueid  : chr "PLT_CN"
##  $ adjtree    : logi FALSE
##  $ prednames  : chr [1:5] "dem" "slp" "asp_cos" "asp_sin" ...
##  $ predfac    : chr "fornf"

Now that we’ve created our population dataset, we can move on to estimation.

modMAarea

Example 2: Area of forest land, Wyoming, 2011-2013

View Example

In this example, we look at estimating the area of forest land in Wyoming from 2011 to 2013 summed to the population unit (sumunit = TRUE) with the generalized regression estimator (MAmethod = "greg"). FIESTA returns raw data for area of forest land, Wyoming, 2011-2013 (sum estimation units).

area1 <- modMAarea(
  MApopdat = MApopdat, # pop - population calculations for WY, post-stratification
  MAmethod = "greg", # est - model-assisted method
  landarea = "FOREST" # est - forest land filter
  )

We can look at the structure of this output with str and the estimates below. Note that again FIESTA outputs a list.

str(area1, max.level = 2)
output
## List of 4
##  $ est    :Classes 'data.table' and 'data.frame':    1 obs. of  3 variables:
##   ..$ ONEUNIT               : num 1
##   ..$ Estimate              : num 660395
##   ..$ Percent Sampling Error: num 9.42
##   ..- attr(*, ".internal.selfref")=<externalptr> 
##   ..- attr(*, "sorted")= chr "ONEUNIT"
##  $ raw    :List of 9
##   ..$ unit_totest  :'data.frame':    1 obs. of  18 variables:
##   ..$ domdat       :'data.frame':    66 obs. of  19 variables:
##   ..$ module       : chr "MA"
##   ..$ esttype      : chr "AREA"
##   ..$ MAmethod     : chr "greg"
##   ..$ predselectlst:List of 1
##   ..$ rowvar       : chr "TOTAL"
##   ..$ colvar       : chr "NONE"
##   ..$ areaunits    : chr "acres"
##  $ statecd: int 56
##  $ invyr  : int [1:3] 2011 2012 2013
area1$est
output
##    ONEUNIT Estimate Percent Sampling Error
## 1:       1 660395.1                   9.42

Example 3: Area of forest land, Wyoming, 2011-2013, using the Elastic Net for variable selection

View Example

Here, we fit the same model as the above example, but rather than using "greg" are our model-assisted method, we can use "gregEN" where the EN stands for “elastic net”. The elastic net performs variable selection for us, grabbing predictors it finds to be most useful in the model.

area2 <- modMAarea(
  MApopdat = MApopdat, # pop - population calculations for WY, post-stratification
  MAmethod = "gregEN", # est - model-assisted method
  landarea = "FOREST", # est - forest land filter
  )

We can again see that the structure of the list is very similar to that in the above example:

str(area2, max.level = 2)
output
## List of 4
##  $ est    :Classes 'data.table' and 'data.frame':    1 obs. of  3 variables:
##   ..$ ONEUNIT               : num 1
##   ..$ Estimate              : num 669219
##   ..$ Percent Sampling Error: num 9.36
##   ..- attr(*, ".internal.selfref")=<externalptr> 
##   ..- attr(*, "sorted")= chr "ONEUNIT"
##  $ raw    :List of 9
##   ..$ unit_totest  :'data.frame':    1 obs. of  18 variables:
##   ..$ domdat       :'data.frame':    66 obs. of  19 variables:
##   ..$ module       : chr "MA"
##   ..$ esttype      : chr "AREA"
##   ..$ MAmethod     : chr "gregEN"
##   ..$ predselectlst:List of 1
##   ..$ rowvar       : chr "TOTAL"
##   ..$ colvar       : chr "NONE"
##   ..$ areaunits    : chr "acres"
##  $ statecd: int 56
##  $ invyr  : int [1:3] 2011 2012 2013

However now the raw list has an item call predselectlst. We can look at this item now:

area2$raw$predselectlst$totest
output
##    ONEUNIT TOTAL           dem         slp     asp_cos asp_sin     fornf2
## 1:       1     1 -0.0001558203 0.008902244 0.001901744       0 -0.3313874

Notably, we can see that dem, slp, asp_cos, and asp_sin were removed from the model.

Example 4: Area by forest type on forest land, Wyoming, 2011-2013

View Example

In this example, we look at adding rows to the output and include returntitle=TRUE to return title information.

area3 <- modMAarea(
    MApopdat = MApopdat,         # pop - population calculations for WY, post-stratification
    MAmethod = "greg",           # est - model-assisted method
    landarea = "FOREST",         # est - forest land filter
    rowvar = "FORTYPCD",         # est - row domain
    returntitle = TRUE           # out - return title information
    )

Again, we can look at the contents of the output list. The output now includes titlelst, a list of associated titles.

str(area3, max.level = 1)
output
## List of 5
##  $ est     :'data.frame':    8 obs. of  3 variables:
##  $ titlelst:List of 9
##  $ raw     :List of 10
##  $ statecd : int 56
##  $ invyr   : int [1:3] 2011 2012 2013

And the estimates:

## Estimate and percent sampling error of estimate
area3$est
output
##   Forest type Estimate Percent Sampling Error
## 1         201  42653.6                  60.56
## 2         265  47917.2                  67.78
## 3         266  71039.9                  43.05
## 4         268  39520.3                   68.8
## 5         281 407596.6                  16.05
## 6         901  24905.9                  78.03
## 7         999  26761.6                  63.54
## 8       Total 660395.1                   9.42

Along with raw data and titles:

## Raw data (list object) for estimate
raw3 <- area3$raw      # extract raw data list object from output
names(raw3)
output
##  [1] "unit_totest"   "unit_rowest"   "domdat"        "module"       
##  [5] "esttype"       "MAmethod"      "predselectlst" "rowvar"       
##  [9] "colvar"        "areaunits"
head(raw3$unit_totest) # estimates by estimation unit (i.e., ESTN_UNIT)
output
##   ONEUNIT      nhat    nhat.var NBRPLT NBRPLT.gt0 ACRES_GIS AREAUSED      est
## 1       1 0.5936603 0.003129767     56         37   1112412  1112412 660395.1
##      est.var   est.se     est.cv      pse CI99left CI99right CI95left CI95right
## 1 3872964796 62233.15 0.09423624 9.423624 500093.1    820697 538420.4  782369.8
##   CI68left CI68right
## 1 598506.8  722283.3
raw3$totest            # estimates for population (i.e., WY)
output
## NULL
head(raw3$unit_rowest) # estimates by row, by estimation unit (i.e., ESTN_UNIT)
output
##   ONEUNIT Forest type       nhat     nhat.var NBRPLT NBRPLT.gt0 ACRES_GIS
## 1       1         201 0.03834336 0.0005392280     56          2   1112412
## 2       1         265 0.04307505 0.0008523855     56          4   1112412
## 3       1         266 0.06386110 0.0007557639     56          5   1112412
## 4       1         268 0.03552668 0.0005973938     56          3   1112412
## 5       1         281 0.36640782 0.0034566717     56         23   1112412
## 6       1         901 0.02238906 0.0003052203     56          1   1112412
##   AREAUSED       est    est.var   est.se    est.cv      pse CI99left CI99right
## 1  1112412  42653.63  667273775 25831.64 0.6056142 60.56142      0.0 109191.53
## 2  1112412  47917.21 1054793996 32477.59 0.6777855 67.77855      0.0 131573.95
## 3  1112412  71039.88  935228490 30581.51 0.4304836 43.04836      0.0 149812.62
## 4  1112412  39520.32  739251639 27189.18 0.6879799 68.79799      0.0 109555.01
## 5  1112412 407596.59 4277497216 65402.58 0.1604591 16.04591 239130.7 576062.46
## 6  1112412  24905.87  377698351 19434.46 0.7803167 78.03167      0.0  74965.72
##    CI95left CI95right   CI68left CI68right
## 1      0.00  93282.72  16965.145  68342.11
## 2      0.00 111572.13  15619.617  80214.81
## 3  11101.23 130978.53  40627.860 101451.90
## 4      0.00  92810.14  12481.820  66558.81
## 5 279409.89 535783.28 342556.478 472636.70
## 6      0.00  62996.71   5579.111  44232.62
head(raw3$rowest)      # estimates by row for population (i.e., WY)
output
## NULL
## Titles (list object) for estimate
titlelst3 <- area3$titlelst
names(titlelst3)
output
## [1] "title.estpse"  "title.unitvar" "title.ref"     "outfn.estpse" 
## [5] "outfn.rawdat"  "outfn.param"   "title.rowvar"  "title.row"    
## [9] "title.unitsn"
titlelst3
output
## $title.estpse
## [1] "Area, in acres, and percent sampling error on forest land by forest type"
## 
## $title.unitvar
## [1] "ONEUNIT"
## 
## $title.ref
## [1] ", 2011-2013"
## 
## $outfn.estpse
## [1] "area_FORTYPCD_forestland"
## 
## $outfn.rawdat
## [1] "area_FORTYPCD_forestland_rawdata"
## 
## $outfn.param
## [1] "area_FORTYPCD_forestland_parameters"
## 
## $title.rowvar
## [1] "Forest type"
## 
## $title.row
## [1] "Area, in acres, on forest land by forest type; , 2011-2013"
## 
## $title.unitsn
## [1] "acres"

Example 5: Area by forest type and stand-size class on forest land, Wyoming, 2011-2013

View Example

In this example, we look at adding rows and columns to output, including FIA names. We also output estimates and percent standard error in the same cell with the allin1 argument in table_options and save data to an outfolder with the outfolder argument in savedata_options.

## Area of forest land by forest type and stand-size class, Wyoming, 2011-2013
area4 <- modMAarea(
    MApopdat = MApopdat,         # pop - population calculations for WY, post-stratification
    MAmethod = "greg",           # est - model-assisted method
    landarea = "FOREST",         # est - forest land filter
    rowvar = "FORTYPCD",         # est - row domain
    colvar = "STDSZCD",          # est - column domain
    savedata = TRUE,             # out - save data to outfolder
    returntitle = TRUE,          # out - return title information
    table_opts = list(
      row.FIAname = TRUE,          # table - row domain names
      col.FIAname = TRUE,          # table - column domain names
      allin1 = TRUE                # table - return output with est(pse)
      ),
    savedata_opts = list(
      outfolder = outfolder,       # save - outfolder for saving data
      outfn.pre = "WY"             # save - prefix for output files
      )
    )

area4$est
output
##                         Forest type     Large diameter    Medium diameter
## 1:                      Douglas-fir  23,330.7 ( 79.00)        -- (    --)
## 2:                 Engelmann spruce  36,892.1 ( 74.16)  11,025.2 (175.15)
## 3: Engelmann spruce / subalpine fir  40,287.9 ( 68.99)  14,443.2 ( 66.30)
## 4:                    Subalpine fir  26,255.3 ( 75.29)        -- (    --)
## 5:                   Lodgepole pine 140,798.2 ( 33.20) 230,288.3 ( 20.42)
## 6:                            Aspen        -- (    --)        -- (    --)
## 7:                       Nonstocked        -- (    --)        -- (    --)
## 8:                            Total 267,564.1 ( 21.17) 255,756.7 ( 19.38)
##        Small diameter        Nonstocked              Total
## 1:  19,322.9 ( 99.12)       -- (    --)  42,653.6 ( 60.56)
## 2:        -- (    --)       -- (    --)  47,917.2 ( 67.78)
## 3:  16,308.7 ( 71.46)       -- (    --)  71,039.9 ( 43.05)
## 4:  13,265.0 (141.72)       -- (    --)  39,520.3 ( 68.80)
## 5:  36,510.1 ( 74.46)       -- (    --) 407,596.6 ( 16.05)
## 6:  24,905.9 ( 78.03)       -- (    --)  24,905.9 ( 78.03)
## 7:        -- (    --) 26,761.6 ( 63.54)  26,761.6 ( 63.54)
## 8: 110,312.6 ( 37.41) 26,761.6 ( 63.54) 660,395.1 (  9.42)

We can again look at the output list, estimates, raw data, and titles:

## Look at output list
names(area4)
output
## [1] "est"      "pse"      "titlelst" "raw"      "statecd"  "invyr"
## Estimate and percent sampling error of estimate
head(area4$est)
output
##                         Forest type     Large diameter    Medium diameter
## 1:                      Douglas-fir  23,330.7 ( 79.00)        -- (    --)
## 2:                 Engelmann spruce  36,892.1 ( 74.16)  11,025.2 (175.15)
## 3: Engelmann spruce / subalpine fir  40,287.9 ( 68.99)  14,443.2 ( 66.30)
## 4:                    Subalpine fir  26,255.3 ( 75.29)        -- (    --)
## 5:                   Lodgepole pine 140,798.2 ( 33.20) 230,288.3 ( 20.42)
## 6:                            Aspen        -- (    --)        -- (    --)
##       Small diameter  Nonstocked              Total
## 1: 19,322.9 ( 99.12) -- (    --)  42,653.6 ( 60.56)
## 2:       -- (    --) -- (    --)  47,917.2 ( 67.78)
## 3: 16,308.7 ( 71.46) -- (    --)  71,039.9 ( 43.05)
## 4: 13,265.0 (141.72) -- (    --)  39,520.3 ( 68.80)
## 5: 36,510.1 ( 74.46) -- (    --) 407,596.6 ( 16.05)
## 6: 24,905.9 ( 78.03) -- (    --)  24,905.9 ( 78.03)
## Raw data (list object) for estimate
raw4 <- area4$raw      # extract raw data list object from output
names(raw4)
output
##  [1] "unit_totest"   "unit_rowest"   "unit_colest"   "unit_grpest"  
##  [5] "domdat"        "module"        "esttype"       "MAmethod"     
##  [9] "predselectlst" "rowvar"        "colvar"        "areaunits"
head(raw4$unit_totest) # estimates by estimation unit (i.e., ESTN_UNIT)
output
##   ONEUNIT      nhat    nhat.var NBRPLT NBRPLT.gt0 ACRES_GIS AREAUSED      est
## 1       1 0.5936603 0.003129767     56         37   1112412  1112412 660395.1
##      est.var   est.se     est.cv      pse CI99left CI99right CI95left CI95right
## 1 3872964796 62233.15 0.09423624 9.423624 500093.1    820697 538420.4  782369.8
##   CI68left CI68right
## 1 598506.8  722283.3
head(raw4$totest)      # estimates for population (i.e., WY)
output
## NULL
head(raw4$unit_rowest) # estimates by row, by estimation unit (i.e., ESTN_UNIT)
output
##   ONEUNIT                      Forest type       nhat     nhat.var NBRPLT
## 1       1                      Douglas-fir 0.03834336 0.0005392280     56
## 2       1                 Engelmann spruce 0.04307505 0.0008523855     56
## 3       1 Engelmann spruce / subalpine fir 0.06386110 0.0007557639     56
## 4       1                    Subalpine fir 0.03552668 0.0005973938     56
## 5       1                   Lodgepole pine 0.36640782 0.0034566717     56
## 6       1                            Aspen 0.02238906 0.0003052203     56
##   NBRPLT.gt0 FORTYPCD ACRES_GIS AREAUSED       est    est.var   est.se
## 1          2      201   1112412  1112412  42653.63  667273775 25831.64
## 2          4      265   1112412  1112412  47917.21 1054793996 32477.59
## 3          5      266   1112412  1112412  71039.88  935228490 30581.51
## 4          3      268   1112412  1112412  39520.32  739251639 27189.18
## 5         23      281   1112412  1112412 407596.59 4277497216 65402.58
## 6          1      901   1112412  1112412  24905.87  377698351 19434.46
##      est.cv      pse CI99left CI99right  CI95left CI95right   CI68left
## 1 0.6056142 60.56142      0.0 109191.53      0.00  93282.72  16965.145
## 2 0.6777855 67.77855      0.0 131573.95      0.00 111572.13  15619.617
## 3 0.4304836 43.04836      0.0 149812.62  11101.23 130978.53  40627.860
## 4 0.6879799 68.79799      0.0 109555.01      0.00  92810.14  12481.820
## 5 0.1604591 16.04591 239130.7 576062.46 279409.89 535783.28 342556.478
## 6 0.7803167 78.03167      0.0  74965.72      0.00  62996.71   5579.111
##   CI68right
## 1  68342.11
## 2  80214.81
## 3 101451.90
## 4  66558.81
## 5 472636.70
## 6  44232.62
head(raw4$rowest)      # estimates by row for population (i.e., WY)
output
## NULL
head(raw4$unit_colest) # estimates by column, by estimation unit (i.e., ESTN_UNIT)
output
##   ONEUNIT Stand-size class       nhat     nhat.var NBRPLT NBRPLT.gt0 STDSZCD
## 1       1   Large diameter 0.24052605 0.0025935705     56         18       1
## 2       1  Medium diameter 0.22991178 0.0019845421     56         14       2
## 3       1   Small diameter 0.09916524 0.0013761149     56          6       3
## 4       1       Nonstocked 0.02405726 0.0002336838     56          1       5
##   ACRES_GIS AREAUSED       est    est.var   est.se    est.cv      pse  CI99left
## 1   1112412  1112412 267564.15 3209443017 56651.95 0.2117322 21.17322 121638.40
## 2   1112412  1112412 255756.70 2455793975 49555.97 0.1937621 19.37621 128108.99
## 3   1112412  1112412 110312.64 1702888801 41266.07 0.3740829 37.40829   4018.28
## 4   1112412  1112412  26761.59  289174599 17005.13 0.6354306 63.54306      0.00
##   CI99right  CI95left CI95right   CI68left CI68right
## 1 413489.89 156528.37 378599.92 211226.172 323902.12
## 2 383404.42 158628.79 352884.62 206475.381 305038.03
## 3 216607.00  29432.62 191192.66  69275.269 151350.01
## 4  70563.91      0.00  60091.04   9850.701  43672.48
head(raw4$colest)      # estimates by column for population (i.e., WY)
output
## NULL
head(raw4$unit_grpest) # estimates by row and column, by estimation unit (i.e., ESTN_UNIT)
output
##   ONEUNIT                                           grpvar        nhat
## 1       1                       Douglas-fir#Large diameter 0.020973074
## 2       1                       Douglas-fir#Small diameter 0.017370285
## 3       1                  Engelmann spruce#Large diameter 0.033164010
## 4       1                 Engelmann spruce#Medium diameter 0.009911038
## 5       1  Engelmann spruce / subalpine fir#Large diameter 0.036216716
## 6       1 Engelmann spruce / subalpine fir#Medium diameter 0.012983698
##       nhat.var NBRPLT NBRPLT.gt0                      Forest type
## 1 2.745298e-04     56          1                      Douglas-fir
## 2 2.964436e-04     56          1                      Douglas-fir
## 3 6.048455e-04     56          3                 Engelmann spruce
## 4 3.013453e-04     56          1                 Engelmann spruce
## 5 6.243626e-04     56          3 Engelmann spruce / subalpine fir
## 6 7.410764e-05     56          1 Engelmann spruce / subalpine fir
##   Stand-size class STDSZCD FORTYPCD ACRES_GIS AREAUSED      est   est.var
## 1   Large diameter       1      201   1112412  1112412 23330.71 339719955
## 2   Small diameter       3      201   1112412  1112412 19322.92 366837446
## 3   Large diameter       1      265   1112412  1112412 36892.05 748472910
## 4  Medium diameter       2      265   1112412  1112412 11025.16 372903083
## 5   Large diameter       1      266   1112412  1112412 40287.92 772624533
## 6  Medium diameter       2      266   1112412  1112412 14443.23  91705337
##     est.se    est.cv       pse CI99left CI99right CI95left CI95right   CI68left
## 1 18431.49 0.7900101  79.00101        0  70807.09        0  59455.77  5001.3621
## 2 19153.00 0.9912064  99.12064        0  68657.78        0  56862.11   276.0668
## 3 27358.23 0.7415752  74.15752        0 107362.19        0  90513.21  9685.4439
## 4 19310.70 1.7515118 175.15118        0  60766.22        0  48873.43     0.0000
## 5 27796.12 0.6899369  68.99369        0 111885.99        0  94767.33 12645.8475
## 6  9576.29 0.6630299  66.30299        0  39110.12        0  33212.41  4920.0086
##   CI68right
## 1  41660.05
## 2  38369.77
## 3  64098.66
## 4  30228.84
## 5  67930.00
## 6  23966.44
head(raw4$grpest)      # estimates by row and column for population (i.e., WY)
output
## NULL
## Titles (list object) for estimate
titlelst4 <- area4$titlelst
names(titlelst4)
output
##  [1] "title.estpse"  "title.unitvar" "title.ref"     "outfn.estpse" 
##  [5] "outfn.rawdat"  "outfn.param"   "title.rowvar"  "title.row"    
##  [9] "title.colvar"  "title.col"     "title.unitsn"
titlelst4
output
## $title.estpse
## [1] "Area, in acres (percent sampling error), by forest type and stand-size class on forest land"
## 
## $title.unitvar
## [1] "ONEUNIT"
## 
## $title.ref
## [1] ", 2011-2013"
## 
## $outfn.estpse
## [1] "WY_area_FORTYPNM_STDSZNM_forestland"
## 
## $outfn.rawdat
## [1] "WY_area_FORTYPNM_STDSZNM_forestland_rawdata"
## 
## $outfn.param
## [1] "WY_area_FORTYPNM_STDSZNM_forestland_parameters"
## 
## $title.rowvar
## [1] "Forest type"
## 
## $title.row
## [1] "Area, in acres (percent sampling error), by forest type on forest land; , 2011-2013"
## 
## $title.colvar
## [1] "Stand-size class"
## 
## $title.col
## [1] "Area, in acres (percent sampling error), by stand-size class on forest land; , 2011-2013"
## 
## $title.unitsn
## [1] "acres"
## List output files in outfolder
list.files(outfolder, pattern = "WY_area")
output
## [1] "WY_area_FORTYPNM_STDSZNM_forestland.csv"                
## [2] "WY_area_FORTYPNM_STDSZNM_forestland_modMA_mase_greg.csv"
list.files(paste0(outfolder, "/rawdata"), pattern = "WY_area")
output
##  [1] "WY_area_FORTYPNM_STDSZNM_forestland_rawdata_colest.csv"                     
##  [2] "WY_area_FORTYPNM_STDSZNM_forestland_rawdata_domdat.csv"                     
##  [3] "WY_area_FORTYPNM_STDSZNM_forestland_rawdata_grpest.csv"                     
##  [4] "WY_area_FORTYPNM_STDSZNM_forestland_rawdata_modMA_mase_greg_domdat.csv"     
##  [5] "WY_area_FORTYPNM_STDSZNM_forestland_rawdata_modMA_mase_greg_unit_colest.csv"
##  [6] "WY_area_FORTYPNM_STDSZNM_forestland_rawdata_modMA_mase_greg_unit_grpest.csv"
##  [7] "WY_area_FORTYPNM_STDSZNM_forestland_rawdata_modMA_mase_greg_unit_rowest.csv"
##  [8] "WY_area_FORTYPNM_STDSZNM_forestland_rawdata_modMA_mase_greg_unit_totest.csv"
##  [9] "WY_area_FORTYPNM_STDSZNM_forestland_rawdata_rowest.csv"                     
## [10] "WY_area_FORTYPNM_STDSZNM_forestland_rawdata_totest.csv"                     
## [11] "WY_area_FORTYPNM_STDSZNM_forestland_rawdata_unit_colest.csv"                
## [12] "WY_area_FORTYPNM_STDSZNM_forestland_rawdata_unit_grpest.csv"                
## [13] "WY_area_FORTYPNM_STDSZNM_forestland_rawdata_unit_rowest.csv"                
## [14] "WY_area_FORTYPNM_STDSZNM_forestland_rawdata_unit_totest.csv"

modMAtree

We will set our estimate variable and filter now. We set estvar to "VOLCFNET" for net cubic foot volume, and filter with estvar.filter set to "STATUSCD == 1" so we only consider live trees in our estimation.

estvar <- "VOLCFNET"
live_trees <- "STATUSCD == 1"

Example 6: Net cubic-foot volume of live trees, Wyoming, 2011-2013

View Example

We now will generate estimates by estimation unit (i.e., ESTN_UNIT) and sum to population (i.e., WY) with modMAtree.

## Return raw data and titles
## Total net cubic-foot volume of live trees (at least 5 inches diameter), Wyoming, 2011-2013 
tree1 <- modMAtree(
    MApopdat = MApopdat,         # pop - population calculations
    MAmethod = "greg",           # est - model-assisted method
    landarea = "FOREST",         # est - forest land filter
    estvar = estvar,             # est - net cubic-foot volume
    estvar.filter = live_trees,  # est - live trees only
    returntitle = TRUE           # out - return title information
    )

names(tree1)
output
## [1] "est"      "titlelst" "raw"      "statecd"  "invyr"
tree1$raw$unit_totest
output
##   ONEUNIT     nhat nhat.var NBRPLT NBRPLT.gt0 ACRES_GIS AREAUSED        est
## 1       1 1026.215 23835.98     56         34   1112412  1112412 1141574550
##       est.var    est.se    est.cv      pse  CI99left  CI99right  CI95left
## 1 2.94961e+16 171744283 0.1504451 15.04451 699190593 1583958507 804961940
##    CI95right  CI68left  CI68right
## 1 1478187159 970782094 1312367006

Example 7: Net cubic-foot volume of live trees, Wyoming, 2011-2013, using the Elastic Net for variable selection

View Example

Here, we fit the same model as the above example, but rather than using "greg" are our model-assisted method, we can use "gregEN" where the EN stands for “elastic net”. The elastic net performs variable selection for us, grabbing predictors it finds to be most useful in the model.

## Return raw data and titles
## Total net cubic-foot volume of live trees (at least 5 inches diameter), Wyoming, 2011-2013 
tree2 <- modMAtree(
    MApopdat = MApopdat,         # pop - population calculations
    MAmethod = "gregEN",         # est - model-assisted method
    landarea = "FOREST",         # est - forest land filter
    estvar = estvar,             # est - net cubic-foot volume
    estvar.filter = live_trees,  # est - live trees only
    returntitle = TRUE           # out - return title information
    )

We can again see that the structure of the list is very similar to that in the above example:

str(tree2, max.level = 2)
output
## List of 5
##  $ est     :Classes 'data.table' and 'data.frame':   1 obs. of  3 variables:
##   ..$ ONEUNIT               : num 1
##   ..$ Estimate              : num 1.16e+09
##   ..$ Percent Sampling Error: num 14.9
##   ..- attr(*, ".internal.selfref")=<externalptr> 
##   ..- attr(*, "sorted")= chr "ONEUNIT"
##  $ titlelst:List of 10
##   ..$ title.estpse : chr "Net merchantable bole  wood volume of live trees (timber species at least 5 inch dia), in cubic feet, and perce"| __truncated__
##   ..$ title.yvar   : chr "Net volume, in cubic feet"
##   ..$ title.estvar : chr "Net merchantable bole  wood volume of live trees (timber species at least 5 inch dia)"
##   ..$ title.unitvar: chr "ONEUNIT"
##   ..$ title.ref    : chr "Wyoming, 2011-2013"
##   ..$ outfn.estpse : chr "tree_VOLCFNET_forestland"
##   ..$ outfn.rawdat : chr "tree_VOLCFNET_forestland_rawdata"
##   ..$ outfn.param  : chr "tree_VOLCFNET_forestland_parameters"
##   ..$ title.tot    : chr "Net merchantable bole  wood volume of live trees (timber species at least 5 inch dia), in cubic feet, on forest"| __truncated__
##   ..$ title.unitsn : chr "cubic feet"
##  $ raw     :List of 13
##   ..$ unit_totest  :'data.frame':    1 obs. of  18 variables:
##   ..$ domdat       :'data.frame':    66 obs. of  19 variables:
##   ..$ plotweights  :List of 1
##   ..$ estvar       : chr "VOLCFNET"
##   ..$ estvar.filter: chr "STATUSCD == 1"
##   ..$ module       : chr "MA"
##   ..$ esttype      : chr "TREE"
##   ..$ MAmethod     : chr "gregEN"
##   ..$ predselectlst:List of 1
##   ..$ rowvar       : chr "TOTAL"
##   ..$ colvar       : chr "NONE"
##   ..$ areaunits    : chr "acres"
##   ..$ estunits     : num 1
##  $ statecd : int 56
##  $ invyr   : int [1:3] 2011 2012 2013

However now the raw list has an item call predselectlst. We can look at this item now:

tree2$raw$predselectlst
output
## $totest
##    ONEUNIT TOTAL      dem       slp  asp_cos   asp_sin    fornf2
## 1:       1     1 0.535073 -55.87839 27.81163 -473.1724 -1186.413

Notably, we can see that [INSERT CORRECT PREDS] dem, slp, asp_cos, and asp_sin were removed from the model.

Example 8: Net cubic-foot volume of live trees by forest type, Wyoming, 2011-2013

View Example

This example adds rows to the output for net cubic-foot volume of live trees (at least 5 inches diameter) by forest type, Wyoming, 2011-2013. We also choose to return titles with returntitle = TRUE.

tree3 <- modMAtree(
    MApopdat = MApopdat,         # pop - population calculations
    MAmethod = "greg",           # est - model-assisted method
    landarea = "FOREST",         # est - forest land filter
    estvar = "VOLCFNET",               # est - net cubic-foot volume
    estvar.filter = "STATUSCD == 1",   # est - live trees only
    rowvar = "FORTYPCD",         # est - row domain 
    returntitle = TRUE           # out - return title information
    )

Again, we investigate the output of the returned list:

## Look at output list
names(tree3)
output
## [1] "est"      "titlelst" "raw"      "statecd"  "invyr"
## Estimate and percent sampling error of estimate
tree3$est
output
##   Forest type     Estimate Percent Sampling Error
## 1         201   15435672.3                  63.81
## 2         265  171222876.3                  70.59
## 3         266  118655149.3                  66.72
## 4         268   43001324.8                  72.97
## 5         281  793259527.2                  19.62
## 6         901           --                     --
## 7         999           --                     --
## 8       Total 1141574549.9                  15.04
## Raw data (list object) for estimate
raw3 <- tree3$raw      # extract raw data list object from output
names(raw3)
output
##  [1] "unit_totest"   "unit_rowest"   "domdat"        "plotweights"  
##  [5] "estvar"        "estvar.filter" "module"        "esttype"      
##  [9] "MAmethod"      "predselectlst" "rowvar"        "colvar"       
## [13] "areaunits"     "estunits"
head(raw3$unit_totest)   # estimates by estimation unit (i.e., ESTN_UNIT)
output
##   ONEUNIT     nhat nhat.var NBRPLT NBRPLT.gt0 ACRES_GIS AREAUSED        est
## 1       1 1026.215 23835.98     56         34   1112412  1112412 1141574550
##       est.var    est.se    est.cv      pse  CI99left  CI99right  CI95left
## 1 2.94961e+16 171744283 0.1504451 15.04451 699190593 1583958507 804961940
##    CI95right  CI68left  CI68right
## 1 1478187159 970782094 1312367006
head(raw3$totest)        # estimates for population (i.e., WY)
output
## NULL
head(raw3$unit_rowest)   # estimates by row, by estimation unit (i.e., ESTN_UNIT)
output
##   ONEUNIT Forest type      nhat    nhat.var NBRPLT NBRPLT.gt0 ACRES_GIS
## 1       1         201  13.87585    78.39277     56          2   1112412
## 2       1         265 153.92033 11805.74969     56          4   1112412
## 3       1         266 106.66472  5065.31148     56          5   1112412
## 4       1         268  38.65592   795.74932     56          2   1112412
## 5       1         281 713.09845 19575.13714     56         23   1112412
## 6       1         901   0.00000     0.00000     56          0   1112412
##   AREAUSED       est      est.var    est.se    est.cv      pse  CI99left
## 1  1112412  15435672 9.700801e+13   9849265 0.6380846 63.80846         0
## 2  1112412 171222876 1.460916e+16 120868349 0.7059124 70.59124         0
## 3  1112412 118655149 6.268127e+15  79171501 0.6672403 66.72403         0
## 4  1112412  43001325 9.847089e+14  31380072 0.7297466 72.97466         0
## 5  1112412 793259527 2.422347e+16 155638920 0.1962018 19.62018 392360235
## 6  1112412         0 0.000000e+00         0       NaN      NaN         0
##    CI99right  CI95left  CI95right  CI68left CI68right
## 1   40805697         0   34739876   5640993  25230351
## 2  482559110         0  408120486  51024394 291421358
## 3  322587423         0  273828441  39922426 197387873
## 4  123831035         0  104505137  11795164  74207485
## 5 1194158819 488212849 1098306206 638483176 948035879
## 6          0         0          0         0         0
head(raw3$rowest)        # estimates by row for population (i.e., WY)
output
## NULL
## Titles (list object) for estimate
titlelst3 <- tree3$titlelst
names(titlelst3)
output
##  [1] "title.estpse"  "title.yvar"    "title.estvar"  "title.unitvar"
##  [5] "title.ref"     "outfn.estpse"  "outfn.rawdat"  "outfn.param"  
##  [9] "title.rowvar"  "title.row"     "title.unitsn"
titlelst3
output
## $title.estpse
## [1] "Net merchantable bole  wood volume of live trees (timber species at least 5 inch dia), in cubic feet, and percent sampling error on forest land by forest type"
## 
## $title.yvar
## [1] "Net volume, in cubic feet"
## 
## $title.estvar
## [1] "Net merchantable bole  wood volume of live trees (timber species at least 5 inch dia)"
## 
## $title.unitvar
## [1] "ONEUNIT"
## 
## $title.ref
## [1] "Wyoming, 2011-2013"
## 
## $outfn.estpse
## [1] "tree_VOLCFNET_FORTYPCD_forestland"
## 
## $outfn.rawdat
## [1] "tree_VOLCFNET_FORTYPCD_forestland_rawdata"
## 
## $outfn.param
## [1] "tree_VOLCFNET_FORTYPCD_forestland_parameters"
## 
## $title.rowvar
## [1] "Forest type"
## 
## $title.row
## [1] "Net merchantable bole  wood volume of live trees (timber species at least 5 inch dia), in cubic feet, on forest land by forest type; Wyoming, 2011-2013"
## 
## $title.unitsn
## [1] "cubic feet"

We can also create a simple barplot from the output:

## Create barplot
datBarplot(
      raw3$unit_rowest, 
      xvar = titlelst3$title.rowvar, 
      yvar = "est"
      )
plot

And a fancier barplot:

## Create fancier barplot
datBarplot(
      raw3$unit_rowest, 
      xvar = titlelst3$title.rowvar, 
      yvar = "est",
      errbars = TRUE, 
      sevar = "est.se", 
      main = FIESTAutils::wraptitle(titlelst3$title.row, 75),
      ylabel = titlelst3$title.yvar, 
      divideby = "million"
      )
plot

Example 9: Net cubic-foot volume of live trees by forest type and stand-size class, Wyoming, 2011-2013

View Example

This examples adds rows and columns to the output, including FIA names, for net cubic-foot volume of live trees (at least 5 inches diameter) by forest type and stand-size class, Wyoming, 2011-2013. We also use the *_options functions to return output with estimates (est) and percent standard error (pse) in same cell - est(pse) with allin1 = TRUE and save data to an outfolder with savedata = TRUE and outfolder = outfolder.

tree4 <- modMAtree(
    MApopdat = MApopdat,         # pop - population calculations
    MAmethod = "greg",           # est - model-assisted method
    landarea = "FOREST",         # est - forest land filter
    estvar = "VOLCFNET",               # est - net cubic-foot volume
    estvar.filter = "STATUSCD  == 1",   # est - live trees only
    rowvar = "FORTYPCD",         # est - row domain
    colvar = "STDSZCD",          # est - column domain
    returntitle = TRUE,          # out - return title information
    savedata = TRUE,             # out - save data to outfolder
    table_opts = table_options(
      row.FIAname = TRUE,          # est - row domain names
      col.FIAname = TRUE,          # est - column domain names
      allin1 = TRUE                # out - return output with est(pse)
    ),
    savedata_opts = savedata_options(
      outfolder = outfolder,       # out - outfolder for saving data
      outfn.pre = "WY"             # out - prefix for output files
      )
    )

Again, we investigate the output of the returned list:

## Look at output list from modGBarea()
names(tree4)
output
## [1] "est"      "pse"      "titlelst" "raw"      "statecd"  "invyr"
## Estimate and percent sampling error of estimate
tree4$est
output
##                         Forest type         Large diameter
## 1:                      Douglas-fir  11,901,294.4 ( 79.00)
## 2:                 Engelmann spruce 131,986,088.6 ( 78.83)
## 3: Engelmann spruce / subalpine fir  99,286,096.2 ( 79.52)
## 4:                    Subalpine fir  43,001,324.8 ( 72.97)
## 5:                   Lodgepole pine 392,479,077.5 ( 34.99)
## 6:                            Aspen            -- (    --)
## 7:                       Nonstocked            -- (    --)
## 8:                            Total 678,653,881.6 ( 24.01)
##           Medium diameter        Small diameter  Nonstocked
## 1:            -- (    --)  3,534,377.8 ( 99.12) -- (    --)
## 2:  39,236,787.7 (175.15)           -- (    --) -- (    --)
## 3:  14,393,410.2 ( 66.30)  4,975,642.9 ( 71.46) -- (    --)
## 4:            -- (    --)           -- (    --) -- (    --)
## 5: 394,877,667.0 ( 21.98)  5,902,782.7 ( 76.80) -- (    --)
## 6:            -- (    --)           -- (    --) -- (    --)
## 7:            -- (    --)           -- (    --) -- (    --)
## 8: 448,507,864.9 ( 23.85) 14,412,803.4 ( 42.87) -- (    --)
##                       Total
## 1:    15,435,672.3 ( 63.81)
## 2:   171,222,876.3 ( 70.59)
## 3:   118,655,149.3 ( 66.72)
## 4:    43,001,324.8 ( 72.97)
## 5:   793,259,527.2 ( 19.62)
## 6:              -- (    --)
## 7:              -- (    --)
## 8: 1,141,574,549.9 ( 15.04)
## Raw data (list object) for estimate
raw4 <- tree4$raw      # extract raw data list object from output
names(raw4)
output
##  [1] "unit_totest"   "unit_rowest"   "unit_colest"   "unit_grpest"  
##  [5] "domdat"        "plotweights"   "estvar"        "estvar.filter"
##  [9] "module"        "esttype"       "MAmethod"      "predselectlst"
## [13] "rowvar"        "colvar"        "areaunits"     "estunits"
head(raw4$unit_totest)   # estimates by estimation unit (i.e., ESTN_UNIT)
output
##   ONEUNIT     nhat nhat.var NBRPLT NBRPLT.gt0 ACRES_GIS AREAUSED        est
## 1       1 1026.215 23835.98     56         34   1112412  1112412 1141574550
##       est.var    est.se    est.cv      pse  CI99left  CI99right  CI95left
## 1 2.94961e+16 171744283 0.1504451 15.04451 699190593 1583958507 804961940
##    CI95right  CI68left  CI68right
## 1 1478187159 970782094 1312367006
head(raw4$totest)        # estimates for population (i.e., WY)
output
## NULL
head(raw4$unit_rowest)   # estimates by row, by estimation unit (i.e., ESTN_UNIT)
output
##   ONEUNIT                      Forest type      nhat    nhat.var NBRPLT
## 1       1                      Douglas-fir  13.87585    78.39277     56
## 2       1                 Engelmann spruce 153.92033 11805.74969     56
## 3       1 Engelmann spruce / subalpine fir 106.66472  5065.31148     56
## 4       1                    Subalpine fir  38.65592   795.74932     56
## 5       1                   Lodgepole pine 713.09845 19575.13714     56
## 6       1                            Aspen   0.00000     0.00000     56
##   NBRPLT.gt0 FORTYPCD ACRES_GIS AREAUSED       est      est.var    est.se
## 1          2      201   1112412  1112412  15435672 9.700801e+13   9849265
## 2          4      265   1112412  1112412 171222876 1.460916e+16 120868349
## 3          5      266   1112412  1112412 118655149 6.268127e+15  79171501
## 4          2      268   1112412  1112412  43001325 9.847089e+14  31380072
## 5         23      281   1112412  1112412 793259527 2.422347e+16 155638920
## 6          0      901   1112412  1112412         0 0.000000e+00         0
##      est.cv      pse  CI99left  CI99right  CI95left  CI95right  CI68left
## 1 0.6380846 63.80846         0   40805697         0   34739876   5640993
## 2 0.7059124 70.59124         0  482559110         0  408120486  51024394
## 3 0.6672403 66.72403         0  322587423         0  273828441  39922426
## 4 0.7297466 72.97466         0  123831035         0  104505137  11795164
## 5 0.1962018 19.62018 392360235 1194158819 488212849 1098306206 638483176
## 6       NaN      NaN         0          0         0          0         0
##   CI68right
## 1  25230351
## 2 291421358
## 3 197387873
## 4  74207485
## 5 948035879
## 6         0
head(raw4$rowest)        # estimates by row for population (i.e., WY)
output
## NULL
head(raw4$unit_colest)   # estimates by column, by estimation unit (i.e., ESTN_UNIT)
output
##   ONEUNIT Stand-size class      nhat    nhat.var NBRPLT NBRPLT.gt0 STDSZCD
## 1       1   Large diameter 610.07402 21448.59554     56         18       1
## 2       1  Medium diameter 403.18490  9244.44106     56         14       2
## 3       1   Small diameter  12.95635    30.85557     56          4       3
## 4       1       Nonstocked   0.00000     0.00000     56          0       5
##   ACRES_GIS AREAUSED       est      est.var    est.se    est.cv      pse
## 1   1112412  1112412 678653882 2.654181e+16 162916561 0.2400584 24.00584
## 2   1112412  1112412 448507865 1.143964e+16 106956241 0.2384713 23.84713
## 3   1112412  1112412  14412803 3.818257e+13   6179205 0.4287302 42.87302
## 4   1112412  1112412         0 0.000000e+00         0       NaN      NaN
##    CI99left  CI99right  CI95left CI95right  CI68left CI68right
## 1 259008631 1098299132 359343290 997964473 516640224 840667539
## 2 173006845  724008885 238877484 658138246 342144388 554871342
## 3         0   30329380   2301785  26523822   8267845  20557762
## 4         0          0         0         0         0         0
head(raw4$colest)        # estimates by column for population (i.e., WY)
output
## NULL
head(raw4$unit_grpest)   # estimates by row and column, by estimation unit (i.e., ESTN_UNIT)
output
##   ONEUNIT                                           grpvar       nhat
## 1       1                       Douglas-fir#Large diameter  10.698636
## 2       1                       Douglas-fir#Small diameter   3.177219
## 3       1                  Engelmann spruce#Large diameter 118.648528
## 4       1                 Engelmann spruce#Medium diameter  35.271801
## 5       1  Engelmann spruce / subalpine fir#Large diameter  89.252961
## 6       1 Engelmann spruce / subalpine fir#Medium diameter  12.938916
##      nhat.var NBRPLT NBRPLT.gt0                      Forest type
## 1   71.436815     56          1                      Douglas-fir
## 2    9.917964     56          1                      Douglas-fir
## 3 8748.214215     56          3                 Engelmann spruce
## 4 3816.641843     56          1                 Engelmann spruce
## 5 5037.268404     56          3 Engelmann spruce / subalpine fir
## 6   73.597319     56          1 Engelmann spruce / subalpine fir
##   Stand-size class STDSZCD FORTYPCD ACRES_GIS AREAUSED       est      est.var
## 1   Large diameter       1      201   1112412  1112412  11901294 8.840029e+13
## 2   Small diameter       3      201   1112412  1112412   3534378 1.227310e+13
## 3   Large diameter       1      265   1112412  1112412 131986089 1.082558e+16
## 4  Medium diameter       2      265   1112412  1112412  39236788 4.722946e+15
## 5   Large diameter       1      266   1112412  1112412  99286096 6.233424e+15
## 6  Medium diameter       2      266   1112412  1112412  14393410 9.107383e+13
##      est.se    est.cv       pse CI99left CI99right CI95left CI95right
## 1   9402143 0.7900101  79.00101        0  36119609        0  30329156
## 2   3503298 0.9912064  99.12064        0  12558275        0  10400715
## 3 104046028 0.7883106  78.83106        0 399990898        0 335912557
## 4  68723696 1.7515118 175.15118        0 216257297        0 173932756
## 5  78952039 0.7951973  79.51973        0 302653071        0 254029249
## 6   9543261 0.6630299  66.30299        0  38975222        0  33097858
##      CI68left CI68right
## 1  2551259.40  21251329
## 2    50495.71   7018260
## 3 28516695.39 235455482
## 4        0.00 107579609
## 5 20771618.84 177800573
## 6  4903039.08  23883781
head(raw4$grpest)        # estimates by row and column for population (i.e., WY)
output
## NULL
## Titles (list object) for estimate
titlelst4 <- tree4$titlelst
names(titlelst4)
output
##  [1] "title.estpse"  "title.yvar"    "title.estvar"  "title.unitvar"
##  [5] "title.ref"     "outfn.estpse"  "outfn.rawdat"  "outfn.param"  
##  [9] "title.rowvar"  "title.row"     "title.colvar"  "title.col"    
## [13] "title.unitsn"
titlelst4
output
## $title.estpse
## [1] "Net merchantable bole  wood volume of live trees (timber species at least 5 inch dia), in cubic feet (percent sampling error), by forest type and stand-size class on forest land"
## 
## $title.yvar
## [1] "Net volume, in cubic feet"
## 
## $title.estvar
## [1] "Net merchantable bole  wood volume of live trees (timber species at least 5 inch dia)"
## 
## $title.unitvar
## [1] "ONEUNIT"
## 
## $title.ref
## [1] "Wyoming, 2011-2013"
## 
## $outfn.estpse
## [1] "WY_tree_VOLCFNET_FORTYPNM_STDSZNM_forestland"
## 
## $outfn.rawdat
## [1] "WY_tree_VOLCFNET_FORTYPNM_STDSZNM_forestland_rawdata"
## 
## $outfn.param
## [1] "WY_tree_VOLCFNET_FORTYPNM_STDSZNM_forestland_parameters"
## 
## $title.rowvar
## [1] "Forest type"
## 
## $title.row
## [1] "Net merchantable bole  wood volume of live trees (timber species at least 5 inch dia), in cubic feet (percent sampling error), by forest type on forest land; Wyoming, 2011-2013"
## 
## $title.colvar
## [1] "Stand-size class"
## 
## $title.col
## [1] "Net merchantable bole  wood volume of live trees (timber species at least 5 inch dia), in cubic feet (percent sampling error), by stand-size class on forest land; Wyoming, 2011-2013"
## 
## $title.unitsn
## [1] "cubic feet"
## List output files in outfolder
list.files(outfolder, pattern = "WY_tree")
output
## [1] "WY_tree_VOLCFNET_FORTYPNM_STDSZNM_forestland.csv"                
## [2] "WY_tree_VOLCFNET_FORTYPNM_STDSZNM_forestland_modMA_mase_greg.csv"
list.files(paste0(outfolder, "/rawdata"), pattern = "WY_tree")
output
##  [1] "WY_tree_VOLCFNET_FORTYPNM_STDSZNM_forestland_rawdata_colest.csv"                     
##  [2] "WY_tree_VOLCFNET_FORTYPNM_STDSZNM_forestland_rawdata_domdat.csv"                     
##  [3] "WY_tree_VOLCFNET_FORTYPNM_STDSZNM_forestland_rawdata_grpest.csv"                     
##  [4] "WY_tree_VOLCFNET_FORTYPNM_STDSZNM_forestland_rawdata_modMA_mase_greg_domdat.csv"     
##  [5] "WY_tree_VOLCFNET_FORTYPNM_STDSZNM_forestland_rawdata_modMA_mase_greg_unit_colest.csv"
##  [6] "WY_tree_VOLCFNET_FORTYPNM_STDSZNM_forestland_rawdata_modMA_mase_greg_unit_grpest.csv"
##  [7] "WY_tree_VOLCFNET_FORTYPNM_STDSZNM_forestland_rawdata_modMA_mase_greg_unit_rowest.csv"
##  [8] "WY_tree_VOLCFNET_FORTYPNM_STDSZNM_forestland_rawdata_modMA_mase_greg_unit_totest.csv"
##  [9] "WY_tree_VOLCFNET_FORTYPNM_STDSZNM_forestland_rawdata_rowest.csv"                     
## [10] "WY_tree_VOLCFNET_FORTYPNM_STDSZNM_forestland_rawdata_totest.csv"                     
## [11] "WY_tree_VOLCFNET_FORTYPNM_STDSZNM_forestland_rawdata_unit_colest.csv"                
## [12] "WY_tree_VOLCFNET_FORTYPNM_STDSZNM_forestland_rawdata_unit_grpest.csv"                
## [13] "WY_tree_VOLCFNET_FORTYPNM_STDSZNM_forestland_rawdata_unit_rowest.csv"                
## [14] "WY_tree_VOLCFNET_FORTYPNM_STDSZNM_forestland_rawdata_unit_totest.csv"

Example 10: Number of live trees by species, Wyoming, 2011-2013

View Example

We can use tree domain in estimation output rows:

## Number of live trees (at least 1 inch diameter) by species
tree5 <- modMAtree(
    MApopdat = MApopdat,         # pop - population calculations
    MAmethod = "greg",           # est - model-assisted method
    landarea = "FOREST",         # est - forest land filter
    estvar = "TPA_UNADJ",               # est - number of trees per acre 
    estvar.filter = "STATUSCD == 1",    # est - live trees only
    rowvar = "SPCD",             # est - row domain
    returntitle = TRUE,          # out - return title information
    table_opts = table_options(    
      row.FIAname = TRUE,          # est - row domain names
      allin1 = FALSE               # out - return output with est and pse
      )
    )

We can also look at the output list and estimates again:

## Look at output list
names(tree5)
output
## [1] "est"      "titlelst" "raw"      "statecd"  "invyr"
## Estimate and percent sampling error of estimate
tree5$est
output
##            Species    Estimate Percent Sampling Error
## 1    quaking aspen  11153476.9                  73.73
## 2      limber pine   4461866.7                  90.07
## 3 Engelmann spruce  50806944.2                  29.77
## 4    subalpine fir  68658860.9                  29.14
## 5   lodgepole pine 184898322.4                  18.59
## 6      Douglas-fir    24655901                  73.84
## 7            Total 344635372.3                  13.07

Example 11: Number of live trees (plus seedlings) by species, Wyoming, 2011-2013

View Example

We can also add seedlings.

Note: seedling data are only available for number of trees (estvar = TPA_UNADJ).

Note: must include seedling data in population data calculations.

MApopdat_seed <- modMApop(popTabs = list(tree = WYtree,
                                         cond = WYcond,
                                         seed = WYseed),
                     pltassgn = WYpltassgn,
                     auxdat = modeldat)
## Number of live trees by species, including seedlings
tree6 <- modMAtree(
    MApopdat = MApopdat_seed,         # pop - population calculations
    MAmethod = "greg",           # est - model-assisted method
    estseed = "add",             # est - add seedling data
    landarea = "FOREST",         # est - forest land filter
    estvar = "TPA_UNADJ",               # est - number of trees per acre 
    estvar.filter = "STATUSCD == 1",    # est - live trees only
    rowvar = "SPCD",             # est - row domain
    returntitle = TRUE,          # out - return title information
    table_opts = table_options(
      row.FIAname = TRUE,          # est - row domain names
      allin1 = FALSE)              # out - return output with est and pse
    )

And again we can look at our outputs and compare estimates:

## Look at output list
names(tree6)
output
## [1] "est"      "titlelst" "raw"      "statecd"  "invyr"
## Estimate and percent sampling error of estimate
tree6$est
output
##            Species    Estimate Percent Sampling Error
## 1    subalpine fir 396440112.5                  23.94
## 2 Engelmann spruce 107965157.9                  33.92
## 3   lodgepole pine 251790363.6                  18.03
## 4      limber pine  38477186.7                  92.27
## 5      Douglas-fir  49218941.8                  80.24
## 6    quaking aspen 105867206.7                  60.84
## 7            Total 949758969.2                  16.07
## Compare estimates with and without seedlings
head(tree5$est)
output
##            Species    Estimate Percent Sampling Error
## 1    quaking aspen  11153476.9                  73.73
## 2      limber pine   4461866.7                  90.07
## 3 Engelmann spruce  50806944.2                  29.77
## 4    subalpine fir  68658860.9                  29.14
## 5   lodgepole pine 184898322.4                  18.59
## 6      Douglas-fir    24655901                  73.84
head(tree6$est)
output
##            Species    Estimate Percent Sampling Error
## 1    subalpine fir 396440112.5                  23.94
## 2 Engelmann spruce 107965157.9                  33.92
## 3   lodgepole pine 251790363.6                  18.03
## 4      limber pine  38477186.7                  92.27
## 5      Douglas-fir  49218941.8                  80.24
## 6    quaking aspen 105867206.7                  60.84