DOC: Update documentation (#159)

* Fix typo in quick start command * Update installation guide * Update overview * Add metric result figure * BUG: fix bad missing years check * Update overview * Update notebooks * Hide in-progress tutorials * Swap out sample BAPs for new BAPs
PEOPLE-ER · Oct 1, 2024 · 65d50ef · 65d50ef
1 parent 0e0a717
commit 65d50ef
Show file tree

Hide file tree

Showing 56 changed files with 152 additions and 296 deletions.
diff --git a/README.md b/README.md
@@ -17,7 +17,7 @@ Github: [https://github.com/PEOPLE-ER/spectral-recovery/](https://github.com/PEO
 Documentation: [https://people-er.github.io/spectral-recovery/](https://people-er.github.io/spectral-recovery/)
 
 PyPi: [https://pypi.org/project/spectral-recovery/](https://pypi.org/project/spectral-recovery/)
-
+    
 ## Overview
 
 `spectral-recovery` is an open-source project and Python package that provides simple, centralized, and reproducible methods for performing [spectral recovery analysis](https://people-er.github.io/spectral-recovery/about/#13-looking-at-recovery-trajectories) to support [Ecosystem Restoration](https://people-er.github.io/spectral-recovery/about/#11-ecosystem-restoration) (ER) efforts in forested ecosystems.
@@ -49,7 +49,7 @@ index_ts = sr.compute_indices(
     indices=["NBR", "NDVI"],
 )
 # Read in restoration site(s)
-rest_site = sr.read_restoration_site(
+rest_site = sr.read_restoration_sites(
     path=data.bc06_wildfire_restoration_site(),
     dist_rest_years={0: [2005, 2006]},
 )

diff --git a/docs/asset/R80P_overview_Figure_1.png b/docs/asset/R80P_overview_Figure_1.png
diff --git a/docs/installation.md b/docs/installation.md
@@ -5,13 +5,7 @@
 spectral_recovery and all of it's required dependencies can be installed using pip (or uv):
 
 ```bash
-pip install spectral_recovery
-```
-
-To install the CLI version of spectral_recovery, use "spectral_recovery[cli]" instead:
-
-```bash
-pip install "spectral_recovery[cli]"
+pip install spectral-recovery
 ```
 
 ## From Source
@@ -23,13 +17,13 @@ pip install "spectral_recovery[cli]"
 spectral_recovery can be installed from source code directly from the project repository:
 
 ```bash
-pip install git+https://github.com/PEOPLE-ER/Spectral-Recovery.git#egg=spectral_recovery
+pip install git+https://github.com/PEOPLE-ER/spectral-recovery.git#egg=spectral_recovery
 ```
 
 or from a local clone of the repository:
 
 ```bash
-git clone https://github.com/PEOPLE-ER/Spectral-Recovery.git
-cd Spectral-Recovery
+git clone https://github.com/PEOPLE-ER/spectral-recovery.git
+cd spectral-recovery
 pip install -e .
 ```
diff --git a/docs/notebooks/MEE_worked_example_32.ipynb b/docs/notebooks/MEE_worked_example_32.ipynb
@@ -214,7 +214,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.11.2"
+   "version": "3.10.10"
   }
  },
  "nbformat": 4,

diff --git a/docs/notebooks/test_data/annual_composites/landsat/2000.tif b/docs/notebooks/test_data/annual_composites/landsat/2000.tif
diff --git a/docs/notebooks/test_data/annual_composites/landsat/2001.tif b/docs/notebooks/test_data/annual_composites/landsat/2001.tif
diff --git a/docs/notebooks/test_data/annual_composites/landsat/2002.tif b/docs/notebooks/test_data/annual_composites/landsat/2002.tif
diff --git a/docs/notebooks/test_data/annual_composites/landsat/2002.tif.aux.xml b/docs/notebooks/test_data/annual_composites/landsat/2002.tif.aux.xml
diff --git a/docs/notebooks/test_data/annual_composites/landsat/2003.tif b/docs/notebooks/test_data/annual_composites/landsat/2003.tif
diff --git a/docs/notebooks/test_data/annual_composites/landsat/2003.tif.aux.xml b/docs/notebooks/test_data/annual_composites/landsat/2003.tif.aux.xml
diff --git a/docs/notebooks/test_data/annual_composites/landsat/2004.tif b/docs/notebooks/test_data/annual_composites/landsat/2004.tif
diff --git a/docs/notebooks/test_data/annual_composites/landsat/2005.tif b/docs/notebooks/test_data/annual_composites/landsat/2005.tif
diff --git a/docs/notebooks/test_data/annual_composites/landsat/2006.tif b/docs/notebooks/test_data/annual_composites/landsat/2006.tif
diff --git a/docs/notebooks/test_data/annual_composites/landsat/2007.tif b/docs/notebooks/test_data/annual_composites/landsat/2007.tif
diff --git a/docs/notebooks/test_data/annual_composites/landsat/2008.tif b/docs/notebooks/test_data/annual_composites/landsat/2008.tif
diff --git a/docs/notebooks/test_data/annual_composites/landsat/2009.tif b/docs/notebooks/test_data/annual_composites/landsat/2009.tif
diff --git a/docs/notebooks/test_data/annual_composites/landsat/2012.tif b/docs/notebooks/test_data/annual_composites/landsat/2012.tif
diff --git a/docs/notebooks/test_data/annual_composites/landsat/2013.tif b/docs/notebooks/test_data/annual_composites/landsat/2013.tif
diff --git a/docs/notebooks/test_data/annual_composites/landsat/2014.tif b/docs/notebooks/test_data/annual_composites/landsat/2014.tif
diff --git a/docs/notebooks/test_data/annual_composites/landsat/2015.tif b/docs/notebooks/test_data/annual_composites/landsat/2015.tif
diff --git a/docs/notebooks/test_data/annual_composites/landsat/2016.tif b/docs/notebooks/test_data/annual_composites/landsat/2016.tif
diff --git a/docs/notebooks/test_data/annual_composites/landsat/2017.tif b/docs/notebooks/test_data/annual_composites/landsat/2017.tif
diff --git a/docs/notebooks/test_data/annual_composites/landsat/2018.tif b/docs/notebooks/test_data/annual_composites/landsat/2018.tif
diff --git a/docs/notebooks/test_data/annual_composites/landsat/2019.tif b/docs/notebooks/test_data/annual_composites/landsat/2019.tif
diff --git a/docs/notebooks/test_data/annual_composites/landsat/2020.tif b/docs/notebooks/test_data/annual_composites/landsat/2020.tif
diff --git a/docs/notebooks/test_data/annual_composites/landsat/2021.tif b/docs/notebooks/test_data/annual_composites/landsat/2021.tif
diff --git a/docs/notebooks/test_data/annual_composites/landsat/2022.tif b/docs/notebooks/test_data/annual_composites/landsat/2022.tif
diff --git a/docs/notebooks/test_data/annual_composites/landsat/2023.tif b/docs/notebooks/test_data/annual_composites/landsat/2023.tif
diff --git a/docs/notebooks/test_data/bc06_wildfire_landsat_bap/2002.tif b/docs/notebooks/test_data/bc06_wildfire_landsat_bap/2002.tif
diff --git a/docs/notebooks/test_data/bc06_wildfire_landsat_bap/2003.tif b/docs/notebooks/test_data/bc06_wildfire_landsat_bap/2003.tif
diff --git a/docs/notebooks/test_data/bc06_wildfire_landsat_bap/2004.tif b/docs/notebooks/test_data/bc06_wildfire_landsat_bap/2004.tif
diff --git a/docs/notebooks/test_data/bc06_wildfire_landsat_bap/2005.tif b/docs/notebooks/test_data/bc06_wildfire_landsat_bap/2005.tif
diff --git a/docs/notebooks/test_data/bc06_wildfire_landsat_bap/2006.tif b/docs/notebooks/test_data/bc06_wildfire_landsat_bap/2006.tif
diff --git a/docs/notebooks/test_data/bc06_wildfire_landsat_bap/2007.tif b/docs/notebooks/test_data/bc06_wildfire_landsat_bap/2007.tif
diff --git a/docs/notebooks/test_data/bc06_wildfire_landsat_bap/2008.tif b/docs/notebooks/test_data/bc06_wildfire_landsat_bap/2008.tif
diff --git a/docs/notebooks/test_data/bc06_wildfire_landsat_bap/2009.tif b/docs/notebooks/test_data/bc06_wildfire_landsat_bap/2009.tif
diff --git a/docs/notebooks/test_data/bc06_wildfire_landsat_bap/2010.tif b/docs/notebooks/test_data/bc06_wildfire_landsat_bap/2010.tif
diff --git a/docs/notebooks/test_data/bc06_wildfire_landsat_bap/2011.tif b/docs/notebooks/test_data/bc06_wildfire_landsat_bap/2011.tif
diff --git a/docs/notebooks/test_data/bc06_wildfire_landsat_bap/2012.tif b/docs/notebooks/test_data/bc06_wildfire_landsat_bap/2012.tif
diff --git a/docs/notebooks/test_data/bc06_wildfire_landsat_bap/2013.tif b/docs/notebooks/test_data/bc06_wildfire_landsat_bap/2013.tif
diff --git a/docs/notebooks/test_data/bc06_wildfire_landsat_bap/2014.tif b/docs/notebooks/test_data/bc06_wildfire_landsat_bap/2014.tif
diff --git a/docs/notebooks/test_data/bc06_wildfire_landsat_bap/2015.tif b/docs/notebooks/test_data/bc06_wildfire_landsat_bap/2015.tif
diff --git a/docs/notebooks/test_data/bc06_wildfire_landsat_bap/2016.tif b/docs/notebooks/test_data/bc06_wildfire_landsat_bap/2016.tif
diff --git a/docs/notebooks/test_data/bc06_wildfire_landsat_bap/2017.tif b/docs/notebooks/test_data/bc06_wildfire_landsat_bap/2017.tif
diff --git a/docs/notebooks/test_data/bc06_wildfire_landsat_bap/2018.tif b/docs/notebooks/test_data/bc06_wildfire_landsat_bap/2018.tif
diff --git a/docs/notebooks/test_data/bc06_wildfire_landsat_bap/2019.tif b/docs/notebooks/test_data/bc06_wildfire_landsat_bap/2019.tif
diff --git a/docs/notebooks/test_data/bc06_wildfire_landsat_bap/2020.tif b/docs/notebooks/test_data/bc06_wildfire_landsat_bap/2020.tif
diff --git a/docs/notebooks/test_data/bc06_wildfire_landsat_bap/2021.tif b/docs/notebooks/test_data/bc06_wildfire_landsat_bap/2021.tif
diff --git a/docs/notebooks/test_data/bc06_wildfire_landsat_bap/2022.tif b/docs/notebooks/test_data/bc06_wildfire_landsat_bap/2022.tif
diff --git a/docs/notebooks/test_data/bc06_wildfire_landsat_bap/2023.tif b/docs/notebooks/test_data/bc06_wildfire_landsat_bap/2023.tif
diff --git a/docs/notebooks/test_data/bc06_wildfire_landsat_bap/2024.tif b/docs/notebooks/test_data/bc06_wildfire_landsat_bap/2024.tif
diff --git a/docs/notebooks/test_data/test_data.zip b/docs/notebooks/test_data/test_data.zip
diff --git a/docs/overview.md b/docs/overview.md
@@ -1,108 +1,204 @@
 # Quick Overview
 
-Here is a simple workflow for computing recovery metrics which demonstrates the core functionalities of the spectral_recovery package. For more detailed examples and in-depth explanations, see the [User Guide](https://PEOPLE-ER.github.io/Spectral-Recovery/terminology).
+This page outlines a simple workflow for computing recovery metrics for a restoration site. For more detailed examples and in-depth explanations, see the [User Guide](https://PEOPLE-ER.github.io/Spectral-Recovery/terminology).
 
-Begin by importing the spectral_recovery tool:
+Begin by importing the spectral-recovery tool:
 
 ``` py
 import spectral_recovery as sr
+from spectral_recovery import data
 ```
+
 ## Define your Restoration Site
 
-The spectral_recovery tool requires a spatial and temporal definition of your restoration site (i.e where is the site and when did the disturbance and restoration events occur?). 
+The spectral-recovery tool requires a spatial and temporal definition of your restoration site (i.e where is the site and when did the disturbance and restoration events occur?). 
+
+Use the `read_restoration_sites` function to create your restoration sites. The function accepts two parameters: a path to a vector file containing site polygons (representing restoration sites) and a dictionary mapping a disturbance and restoration year to each polygon (e.g {0: [2007, 2008], 1: [2002, 2003]}). It returns a geopandas.GeoDataFrame. 
 
-The read_restoration_polygon function can read any vector format file containing your resoration polygon and will return a geopandas.GeoDataFrame containing your polygon and dates:
+Each row in the returned GeoDataframe contains a unique restoration site, defined by a polygon, disturbance year, and restoration year. Multiple restoration sites can be defined at once, each with distinct years. Once read in, every restoration site can be uniquely identified and accessed by the index of the row it is in the Dataframe (i.e row 0 contains restoration site 0, row 1 contains restoration site 1, etc.).
 
-``` py 
-rest_site = sr.read_restoration_polygon(
-    path="my_restoration_poly.gpkg",
-    disturbance_start="2005",
-    restoration_start="2006",
+``` py
+rest_site = sr.read_restoration_sites(
+    path=data.bc06_wildfire_restoration_site(),
+    dist_rest_years={0: [2005, 2006]}
 )
+# Accessing restoration site 0
+rest_site[0]
+#    dist_start  rest_start                                           geometry
+# 0        2005        2006  MULTIPOLYGON (((493414.458 5966949.912, 493414...
 ```
 
 ## Read in Timeseries Data
 
-The spectral_recovery tool also requires an annual timeseries of rasters as input. If your images are written to disk (TIF), spectral_recovery can help you read your images into a xarray.DataArray object with appropriate dimension labels and spatial attributes. By default, the images are read in lazily using dask.arrays. 
+The spectral-recovery tool also requires a timeseries of optical satellite imagery or spectral indices covering the extent of your restoration sites. This is the spectral data from which disturbance, recovery, and reference values will eventually be derived. Each image in the timeseries should represent a unique year and it is strongly reccomened, but not required, that each image be a composite of multiple images to better account for temporal variation. 
+
+If your timeseries of images are written to disk, the `read_timeseries` function can read the images from a directory path or dictionary of file paths into an xarray.DataArray, adding appropriate dimension labels and spatial attributes.
+
+An image can contain spectral bands or spectral indices. If spectral bands are read in, the next section demonstrates how to compute spectral indices from spectral bands. If spectral indices are read in, then continue on with the spectral-recovery workflow.
 
 ``` py
 ts = sr.read_timeseries(
-    path_to_tifs="tifs/",
-    band_names={1: "blue", 2: "green", 3: "red", 4: "nir", 5: "swir16" }
+    path_to_tifs=data.bc06_wildfire_landsat_bap_timeseries(),
+    band_names={1: "blue", 2: "green", 3: "red", 4: "nir", 5: "swir16", 6: "swir22" }
 )
+ts
+# <xarray.DataArray (band: 6, time: 23, y: 34, x: 56)> Size: 2MB
+# dask.array<getitem, shape=(6, 23, 34, 56), dtype=float64, chunksize=(1, 1, 34, 56), chunktype=numpy.ndarray>
+# Coordinates:
+#   * x            (x) float64 448B 4.923e+05 4.923e+05 ... 4.939e+05 4.939e+05
+#   * y            (y) float64 272B 5.968e+06 5.968e+06 ... 5.967e+06 5.967e+06
+#     spatial_ref  int32 4B 0
+#   * time         (time) datetime64[ns] 184B 2006-01-01 2007-01-01 ... 2017-01-01
+#   * band         (band) <U2 48B 'B' 'G' 'R' 'N' 'S1' 'S2'
+# Attributes:
+#     AREA_OR_POINT:  Area
+#     scale_factor:   1.0
+#     add_offset:     0.0
 ```
 
+
 ## Compute Indices
 
-Select a set of indices. The indices you select should represent the characteristics you are interested in computing recovery metrics for.
+Recovery metrics for restoration sites are computed using spectral indices, which can act as proxies for vegetation health and characteristics. If your timeseries is composed of spectral bands, use the `compute_indices` function to compute spectral indices. The spectral indices you choose should depend on which [ecosystem or vegetation characteristics](https://people-er.github.io/spectral-recovery/spectral_recovery_tool/#333-spectral-indices) you are interested in monitoring within your restoration sites.
 
 !!! info
 
     spectral_recovery uses the modern spectral index catalogue [Awesome Spectral Indices](https://awesome-ee-spectral-indices.readthedocs.io/en/latest/) to support computation of 100+ vegetation and burn indices. Since 100+ choices can be overwhelming, a smaller list of "core" indices, indices that have been proven effective in previous spectral recovery research, is provided [here](https://PEOPLE-ER.github.io/Spectral-Recovery/spectral_recovery_tool/#333-spectral-indices). 
 
 ``` py   
 indices = sr.compute_indices(ts, indices=["NBR", "GNDVI", "SAVI"])
+indices
+# <xarray.DataArray (band: 3, time: 23, y: 34, x: 56)> Size: 1MB
+# dask.array<concatenate, shape=(3, 23, 34, 56), dtype=float64, chunksize=(1, 1, 34, 56), chunktype=numpy.ndarray>
+# Coordinates:
+#  * x            (x) float64 448B 4.923e+05 4.923e+05 ... 4.939e+05 4.939e+05
+#  * y            (y) float64 272B 5.968e+06 5.968e+06 ... 5.967e+06 5.967e+06
+#  * time         (time) datetime64[ns] 184B 2002-01-01 2003-01-01 ... 2024-01-01
+#  * band         (band) <U5 60B 'NBR' 'GNDVI' 'SAVI'
+#    spatial_ref  int32 4B 0
 ```
 
 ## Compute Recovery Targets
 
-Some recovery metrics require a recovery target which represents the desired spectral characteristics of your resoration site. A novel feature of the `spectral_recovery` package is that it allows users to choose between using a [historic recovery target](https://people-er.github.io/Spectral-Recovery/about/#14-recovery-targets) or a [reference recovery target](https://people-er.github.io/Spectral-Recovery/about/#14-recovery-targets) when computing recovery metrics. Be sure to consult the user guide and current literature when deciding which type of target works best for computing your recovery metrics.
+Some recovery metrics require a recovery target, which are the desired spectral characteristics of your resoration site. A novel feature of `spectral-recovery` is that it allows users to choose between using a [historic recovery target](https://people-er.github.io/Spectral-Recovery/about/#14-recovery-targets) or a [reference recovery target](https://people-er.github.io/Spectral-Recovery/about/#14-recovery-targets) when computing recovery metrics. Be sure to consult the user guide and current literature when deciding which type of target works best for your restoration needs.
 
 !!! tip
 
-    Only Y2R and R80P require recovery targets. If you are not computing these metrics, you can skip computing a recovery target and not pass anything to the `recovery_target` parameter in `compute_metrics` (default is None).
+    Only Y2R and R80P require recovery targets. If you are not computing these metrics, you can skip computing a recovery target and do not use the `recovery_target` parameter in `compute_metrics` (default is None).
 
-The spectral_recovery tool provides 2 methods to faciliate recovery target computation. For more information on the recovery target methods that are available with the tool, see [Recovery Targets](http://127.0.0.1:8000/Spectral-Recovery/recovery_targets/).
+The spectral_recovery tool provides 3 built-in functions to faciliate recovery target computation. 
 
 ### Historic Recovery Targets
 
-To compute a recovery target based on the historic conditions of your restoration site:
+When computing historic recovery targets, a unique set of recovery targets will be returned for each restoration site passed to the function. The recovery targets are returned in a dictionary of xarray.DataArrays. The keys for the dictionary are the restoration site IDs, e.g if there are two restoration sites (rs) passed, rs 0 and rs 1, the returned dictionary of recovery targets would be {0: <DataArray of rs 0 targets>, 1: <DataArray of rs 1 targets>}.
+
+The first option to compute historic recovery targets is the `targets.historic.median` function. `targets.historic.median` computes the median of historic (pre-disturbance) values within each restoration site.
+
+`targets.historic.median` can compute two difference scales of recovery target, polygon and pixel, passed through the `scale` parameter. `scale="polygon"` will return a single recovery target for the entire restoration site. `scale="pixel` will return a recovery target for every pixel in the restoration site.
 
 ``` py
-hist_targets = sr.targets.median_target(
-    polygon="reference_site_polygons.gpkg"
-    timeseries_data=indices
-    reference_start="2022"
-    reference_end="2022"
+hist_targets_med_poly = sr.targets.historic.median(
+    restoration_sites=rest_site,
+    timeseries_data=indices,
+    reference_years={0: [2003, 2005]},
     scale="polygon"
 )
+hist_targets_med_pix = sr.targets.historic.median(
+    restoration_sites=rest_site,
+    timeseries_data=indices,
+    reference_years={0: [2003, 2005]},
+    scale="pixel"
+)
+hist_targets_med_poly
+# {0: <xarray.DataArray (band: 3)> Size: 24B ... }
+hist_targets_med_pix
+# {0: <xarray.DataArray (band: 3, y: 33, x: 56)> ...}
+```
+
+The second historic recovery target function is `targets.historic.window` which first performs a moving window average to account for spatial variation, then derives historic median recovery targets from the averaged arrays. This function only returns targets at the polygon scale.
+
+``` py
+hist_targets_win = sr.targets.historic.window(
+    restoration_sites=rest_site,
+    timeseries_data=indices,
+    reference_years={0: [2003, 2005]},
+)
+hist_targets_win
+# {0: <xarray.DataArray (band: 3, y: 33, x: 56)> ... }
 ```
 
 ### Reference Recovery Targets
 
-To compute a recovery target based conditions in a reference site, you will need to provide a path to vector file or a geopandas.DataFrame containing your reference site polygon(s) alongside the timeseries data and reference years. 
+To compute a recovery targets based on reference systems, you will need to provide a path to vector file or a geopandas.DataFrame containing your reference site polygon(s). These are locations with ideal or desired characteristics in relation to your restoration site. The `targets.reference.median` function will take the median values across the reference system (site and years) and return a single recovery target value.
 
 ``` py
-ref_targets = sr.targets.median_target(
-    polygon="reference_sys_polygons.gpkg"
-    timeseries_data=indices
-    reference_start="2022"
-    reference_end="2022"
-    scale="polygon"
+ref_targets = sr.targets.reference.median(
+    reference_sites="reference_polygons.gpkg",
+    timeseries_data=indices,
+    reference_start="2022",
+    reference_end="2023",
 )
+ref_targets
+# {0: <xarray.DataArray (band: 3)> ... }
 ```
 
 ## Compute Recovery Metrics
 
-Finally, once you've defined your restration site, computed your indices, and derived your recovery targets, you can compute recovery metrics:
+Finally, once you've defined your restoration site, computed your indices, and derived your recovery targets, you can compute recovery metrics. `compute_metrics` will return recovery metrics for each given restoration site. The recovery metrics are returned in a dictionary of xarray.DataArrays. The keys for the dictionary are the restoration site IDs, e.g if there are two restoration sites (rs) passed, rs 0 and rs 1, the returned dictionary of metrics would be {0: <DataArray of rs 0 metrics>, 1: <DataArray of rs 1 metrics>}.
 
 ``` py
 metrics = sr.compute_metrics(
-    metrics=["dNBR", "Y2R", "R80P"],
-    restoration_polygons=restoration_site,
+    metrics=["Y2R", "R80P", "deltaIR"],
+    restoration_sites=rest_site,
     timeseries_data=indices,
-    recovery_target=hist_targets
+    recovery_targets=hist_targets_med_poly
 )
+metrics
+# {0: <xarray.DataArray (metric: 3, band: 3, y: 33, x: 56)> Size: 133kB
+# dask.array<concatenate, shape=(3, 3, 33, 56), dtype=float64, chunksize=(1, 1, 33, 56), chunktype=numpy.ndarray>
+# Coordinates:
+#   * x            (x) float64 448B 4.923e+05 4.923e+05 ... 4.939e+05 4.939e+05
+#   * y            (y) float64 264B 5.968e+06 5.968e+06 ... 5.967e+06 5.967e+06
+#   * band         (band) <U5 60B 'NBR' 'GNDVI' 'SAVI'
+#   * metric       (metric) <U7 84B 'Y2R' 'R80P' 'deltaIR'
+#     spatial_ref  int32 4B 0}
+```
+
+You can investigate individual metrics by selecting metric name from the metrics dimension
+
+``` py
+# Access restoration site 0's recovery metrics
+rs_0_metrics = metrics[0]
+r80p = rs_0_metrics.sel(metric="R80P").compute()
+r80p
+# <xarray.DataArray (band: 3, y: 33, x: 56)> Size: 44kB
+# array( ... )
+# Coordinates:
+#   * x            (x) float64 448B 4.923e+05 4.923e+05 ... 4.939e+05 4.939e+05
+#   * y            (y) float64 264B 5.968e+06 5.968e+06 ... 5.967e+06 5.967e+06
+#   * band         (band) <U5 60B 'NBR' 'GNDVI' 'SAVI'
+#     metric       <U7 28B 'R80P'
+#     spatial_ref  int32 4B 0
 ```
 
-You can investigate individual metrics by selecting the metric name from the metrics dimension:
+Visualize recovery metrics over your restoration site using Xarray's `imshow` function
 
 ``` py
-y2r = metric.sel(metrics="Y2R")
+import matplotlib.pyplot as plt 
+
+r80p.sel(band="NBR").plot.imshow()
+plt.show()
 ```
 
-or you can write metrics to file using the `to_raster` function in the rioxarray extension:
+![image](asset/R80P_overview_Figure_1.png)
+
+
+or you can write metrics to file using the `rio.to_raster` function in the rioxarray extension:
 
 ``` py
-y2r.rio.to_raster("y2r.tif")
+r80p.rio.to_raster("r80p.tif")
 ```
+
+## More Examples?
+
+For more examples of the spectral-recovery tool, see our [notebooks](https://github.com/PEOPLE-ER/spectral-recovery/tree/main/docs/notebooks) or run them in a [Binder instance](https://mybinder.org/v2/gh/PEOPLE-ER/spectral-recovery/HEAD?labpath=docs%2Fnotebooks%2F)
diff --git a/docs/overview_CLI.ipynb b/docs/overview_CLI.ipynb