to-do (work in progress)

[ErinBecker]

I'm working through the lesson now and making myself a to-do list of things that need clean up or clarification.

All episodes

• Change rasterToPoints calls to as.raster::data.frame syntax. Addressed in More cleanup #223
• Change \n in plot titles to subtitle = . Addressed in More cleanup #223
• Change from HTML to Markdown syntax when appropriate (for stylistic similarity across lessons) Addressed in More cleanup #223

Episode 1

• Explain the output of DSM_HARV - learners should be helped to identify the salient features of this type of output. Addressed with Check early lessons #181
• Break the code block that starts "convert to a df for plotting in two steps," into two chunks, first for the DTM conversion, second for the plotting. Have learners check the structure of the DTM before plotting (good point for exercise). Addressed in more cleanup as suggested in #165 #167
• Use pipes instead of creating intermediate object (DSM_HARV_pts). Learners will have used pipes in a previous lesson. This will cut down two lines of code (no longer need the rm() call). Addressed in more cleanup as suggested in #165 #167
• Remove ggtitle() call in first ggplot() call to cut down on typing.(?) Addressed in more cleanup as suggested in #165 #167
• Question about metadata (immediately after first plot) is out of place. Episode moves directly into talking about CRSs rather than metadata. The connection is explained later on, but it may not be clear to the learner immediately why we ask this question before talking about the CRS. Maybe just mention that the CRS encodes important metadata? Addressed with Check early lessons #181
• Move info introducing CRS to organization lesson. This includes the sections titled "Coordinate Reference Systems, What Makes Spatial Data Line Up On A Map?, and How Map Projections Can Fool the Eye. Addressed in move intro to raster to organization lesson #164
• "crs() method" should be "crs() function" Addressed in more cleanup as suggested in #165 #167
• the following parts of the CRS object aren't explained +no_defs +ellps=WGS84 +towgs84=0,0,0. Maybe link to more information about interpreting a CRS object in PROJ 4 format. Addressed with Cleanup to episode 9 #180
• move "Extent" and "Resolution" sections to organization lesson (with images) Addressed in move intro to raster to organization lesson #164
• remove repeated call to crs() (after Resolution section). Addressed in move intro to raster to organization lesson #164
• simplify "square of rectangular" to "rectangular" Addressed in more cleanup as suggested in #165 #167
• the "NoData Value standard" section mentions that GeoTiffs sometimes contain information about what is being used as the NoData value. The very end of this episode shows how to find that information. Maybe mention that that will be shown later. Addressed with Check early lessons #181
• The "Find Bad Data values" section says "Sometimes a raster’s metadata will tell us the range of expected values for a raster." How would we find this range if it existed for our data? Exercise for learners. Addressed with Check early lessons #181
• remove "xlab" and "ggtitle" calls in ggplot calls (multiple) to minimize typing. (?) Addressed in more cleanup as suggested in #165 #167
• In the sentence "Jump to the fourth tutorial in this series for a tutorial on multi-band rasters: Work with Multi-band Rasters: Images in R." change "the fourth" to "a later" in case episode order is changed at a later point. Addressed with Check early lessons #181
• The last section "View Raster File Attributes" says that it is ideal to view raster attributes before importing into R. We should mention that for the purposes of this tutorial, we did it in the opposite direction because we wanted to be able to explain each piece of the metadata individually. Addressed with Check early lessons #181
• Replace description of proj4 formatting with wording from Episode 9. It is more detailed and clearer. Addressed with Cleanup to episode 9 #180

Episode 2

• Remove repetitive steps in intro. We can assume that learners are completing the episodes in order, so we can remove anything that happened in episode 1. This will also help Maintainers as changes only need to be made in one place, rather that in every location where the information is repeated. Addressed with Implement suggestions from #165 #169
• Remove calls to xlab, ylab, and ggtitle for reasons discussed above. Addressed with Implement suggestions from #165 #169
• Mention ( ] notation in callout. Some learners may not be familiar with the inclusive, exclusive range notation. Addressed with Implement suggestions from #165 #169
• In the "Layering Rasters" section split the code chunks as in the first episode. Addressed with Implement suggestions from #165 #169

Episode 3

• [ ] Remove the steps of reading in the data and converting to dfs as the learners will have done that in previous episodes. See discussion in simplifying which data is used in the first few episodes #170.
• Remove axis labels and titles for reasons discussed above. Simplify plotting calls as much as possible. Addressed with Implement suggestions from #165 #169
• Maybe a little more information about how WGS84 and UTM are different.(?) HELP NEEDED

Episode 4

• Change "tutorial" to "episode" throughout (for consistency in terminology with other Carpentries lessons). Addressed with Implement suggestions from #165 #169
• Learners have already loaded in the data for this episode. Remove that redundancy to avoid confusion. Addressed with Implement suggestions from #165 #169
• Keep GDALinfo calls but ask learners to interpret something from them (challenge) Addressed with Implement suggestions from #165 #169
• Simplify ggplot calls as in other episodes. Addressed with Implement suggestions from #165 #169
• Can the function call within overlay() be simplified? Currently fun=function(r1, r2){return(r1-r2)})
• Move data tip about custom function definitions to before the learners implement overlay for the first time. Addressed with Implement suggestions from #165 #169
• Challenges are long and complex. Can they be broken down into smaller challenges that instructors can pick and choose from?

Episode 5

• Move everything before "Getting Started with Multi-Band Data in R" to the organization lesson. Addressed with Move intro to multiband rasters from raster vector lesson organization-geospatial#31
• Change "tutorial" to "episode" throughout (for consistency in terminology with other Carpentries lessons). Addressed with Cleanup episode 5 #171
• Simplify ggplot calls as in other episodes. Keep titles, but shorten. Addressed with Cleanup episode 5 #171
• Is this the first time learners are using the rgdal package? If so, keep the line of code to load that library, but set echo = FALSE for the other packages learners already have loaded. Addressed with Cleanup episode 5 #171
• Needs to be ggplotized. Discussed in convert static plotting from base r to ggplot2 #133

Episode 6

• Move "About Vector Data" section to organization lesson. Addressed with Move intro to multiband rasters from raster vector lesson organization-geospatial#31
• Learners will already have raster and rgdal loaded. Only need to have them load sf. Addressed with More cleanup #173
• The challenge "Plot Raster & Vector Data Together" seems pretty advanced. Can we break this up into multiple challenge Qs? Maybe first add the roads and then the tower location?
• The learners may not understand the power or importance of plotting vector data - currently only two vector data plots are shown and both are simple polygons (rectangles). It would be powerful to overlay that rectangle (the field site boundaries) over the site map (raster data) before having the learners do the exercise.
• In the exercise solution, the roads extend past the boundaries of the map. Should that be the case? Learners will probably want to know how to restrict the range of the vector layers. Could this be another exercise (optional)?
• Needs to be ggplotized. Discussed in convert static plotting from base r to ggplot2 #133
• Remove "Additional Resources: Plot Parameter Options" section. We are no longer using base plotting. Addressed in More cleanup #173. Note: Episode 7 has a lot about plot customization - line thickness, color selection, legend.
• Does the "Shapefile Attributes" section need to be here? It is repeated in the next episode where it appears to be used more. Addressed in More cleanup #173
• The "Plot Multiple Shapefiles" section is repeated pretty much verbatim at the beginning of Episode 8. We can delete it here and the associated challenge. Addressed in More cleanup #173
• Move the exercise on plotting vector and raster data together to the end of Episode 8, where learners have spent more time making multi-layer plots. Addressed in More cleanup #173

Episode 7

• Move "About Shapefiles Attribtes" section to project organization lesson. Addressed in Cleanup episodes 6, 7 and 8 #179 See also Move info on spatial data attributes from raster vector lesson organization-geospatial#32
• The answers to the first exercise don't match the comments in the answers code. There appear to be 15 and 2 attributes, respectively, not 14 and 1 as shown in the comments. Addressed in Cleanup episodes 6, 7 and 8 #179
• Move comments to sentences in episode text to match style of other DC lessons. Addressed in Cleanup episodes 6, 7 and 8 #179
• The features count is listed incorrectly later in the episode to be 13, not 15. Addressed in Cleanup episodes 6, 7 and 8 #179
• Split the "Subset Spatial Line Objects" exercise into two separate exercises to make it easier to use in the classroom. Addressed in Cleanup episodes 6, 7 and 8 #179
• remove unnecessary lines of code from the " Subset Spatial Line Objects" exercise. Addressed in Cleanup episodes 6, 7 and 8 #179
• In the data tip about as.factor, refer back to the appropriate episode in the intro R lesson. Addressed in Cleanup episodes 6, 7 and 8 #179
• The last two parts of the last exercise are repeated at the end of episode 8 and make more sense there. Remove them here. Addressed in Cleanup episodes 6, 7 and 8 #179

Episode 8

• The "Plot Data" section repeats the end of the previous lesson and so can be removed. Addressed in Cleanup episodes 6, 7 and 8 #179
• Change "Customize your legend" title to "Plot multiple shapefiles"? (we already did some legend customization in the previous episode, but haven't yet plotted multiple shapefiles in a single plot.) Addressed in Cleanup episodes 6, 7 and 8 #179
• The last plotting call saves to an object called gg, but this is the first time we've done this. Make this like the other plot calls. Addressed in Cleanup episodes 6, 7 and 8 #179
• Error in f(...): object 'road_palette' not found

Episode 9

• Hide output from reading in the state boundary file
• Remove comments and move that information to the narrative text to fit the rest of DC lessons. Addressed with Cleanup to episode 9 #180
• Replace "continental" with "contiguous" - "continental" includes Alaska, which is not shown here. Addressed with Cleanup to episode 9 #180
• Move additional resources to a callout box. Addressed with Cleanup to episode 9 #180
• Add reference back to first episode when talking about proj4 format. Addressed with Cleanup to episode 9 #180
• Needs to be ggplotized. Discussed in convert static plotting from base r to ggplot2 #133
• After ggplotizing, make sure challenge and answer text are concise and clear.

Episode 10

• The last objective ("Be able to plot raster and vector data in the same plot to create a map.") is something that they've done already in an earlier lesson. (I think)
• stringsAsFactors should be TRUE. All of the character vectors are categorical data. Addressed in Clean-up to episode 10 and 11 #217
• 21 locations (rows) and 15 variables (attributes). There are 16 variables. Addressed in Clean-up to episode 10 and 11 #217
• Elsewhere in the lesson data.frame is "dataframe". It is only data.frame here and episode 14. Addressed in Clean-up to episode 10 and 11 #217
• Move comments to narrative for stylistic consistency. Addressed in Clean-up to episode 10 and 11 #217
• "View the column names, we can see that our data.frame that contains several fields that might contain spatial information." --> Our column names include several fields that might contain spatial information.
• Remove two lines of code showing that you can access columns by column number. Easier by name and already demonstrated by number in an earlier lesson. Addressed in Clean-up to episode 10 and 11 #217
• Fix link to metadata episode and other episode links. Addressed in Clean-up to episode 10 and 11 #217
• "This link shows the proj4 string for UTM Zone 18N WGS84." can be replaced with an output box. Addressed in Clean-up to episode 10 and 11 #217
• Remove "simply" and "we can see". Addressed in Clean-up to episode 10 and 11 #217
• Move import of lines shapefile and boundary object to file header. Addressed in Clean-up to episode 10 and 11 #217
• Needs to be ggplotized. Discussed in convert static plotting from base r to ggplot2 #133 Addressed in Clean-up to episode 10 and 11 #217
• "When we plot several spatial layers in R, the first layer that is plotted, becomes the extent of the plot. If we add additional layers that are outside of that extent, then the data will not be visible in our plot." -- I know this is true for base plot, but is it true for ggplot? I think that ggplot looks at all the layers when creating the plot canvas. We may be able to delete this entire section. Addressed in Clean-up to episode 10 and 11 #217
• Weird bullet point at "so what is going on?" Addressed in Clean-up to episode 10 and 11 #217

Episode 11

• Add loading in data step to start of file, echo = FALSE Addressed in Clean-up to episode 10 and 11 #217
• Add the data file from csv to shapefile episode to the loading in data step Addressed in Clean-up to episode 10 and 11 #217
• Add "we will continue working with . . . [ files ] " at start of file - stylistic consistency Addressed in Clean-up to episode 10 and 11 #217
• typo "keep reduce file sizes" Addressed in Clean-up to episode 10 and 11 #217
• collapse three one-two line paragraphs to a single para Addressed in Clean-up to episode 10 and 11 #217
• Needs to be ggplotized. Discussed in convert static plotting from base r to ggplot2 #133 Partially addressed in Clean-up to episode 10 and 11 #217 Finished in Cleanup to episode 11 #220
• Move comments to narrative for stylistic consistency. Addressed in Cleanup to episode 11 #220
• Remove line starting "If you completed .csv to Shapefile in R you have these plot" Addressed in Cleanup to episode 11 #220
• Move bullet point extra resources to callout (multiple places) Addressed in Cleanup to episode 11 #220
• Broken image link at start of "Extract Raster Pixels Values Using Vector Polygons" Addressed in Cleanup to episode 11 #220
• Sentence fragment - meant to be a section header? "Extract raster information using a polygon boundary." Addressed in Cleanup to episode 11 #220
• "In this case," used twice in adjoining sentences. Addressed in Cleanup to episode 11 #220
• data.frame and list should be plain text. Addressed in Cleanup to episode 11 #220
• Broken header "##Extract Data using x,y Locations" Addressed in Cleanup to episode 11 #220
• Broken image link in ##Extract Data using x,y Locations section Addressed in Cleanup to episode 11 #220
• Address comments from @obrl-soil in Cleanup to episode 11 #220.

Episode 12

• hide messages/warnings from opening code Addressed in More cleanup #223
• move intro to time series data and image to organization lesson Addressed in Move ndvi intro and general cleanup organization-geospatial#33
• Add pointer here to organization lesson Addressed in More cleanup #223
• "greeneness" is misspelled in the figure title
• Move bullet point extra resources to callout (multiple places) Addressed in More cleanup #223
• Broken image link in NDVI section Addressed in More cleanup #223
• Move NDVI and RGB sections to organization lesson Addressed in Move ndvi intro and general cleanup organization-geospatial#33
• "Make sure you have" these libraries loaded Addressed in More cleanup to episode 12 #224
• Add loading in data step to start of file, echo = FALSE Addressed in More cleanup to episode 12 #224
• Add "we will continue working with . . . [ files ] " at start of file - stylistic consistency Addressed in More cleanup to episode 12 #224
• Explain pattern=".tif$" in callout box and add link to regex cheat sheet Addressed in More cleanup to episode 12 #224
• convert metadata section to exercise (there already is an exercise, but it's repetitive with the code, move code to exercise) Addressed in More cleanup to episode 12 #224
• shrink images Addressed in More cleanup #223
• break large code chunks into smaller chunks and move comments to narrative Addressed in More cleanup to episode 12 #224
• convert "Take a Closer Look at Our Data" to exercise Addressed in More cleanup to episode 12 #224
• Move "Hint" to callout (Julian days) Addressed in More cleanup #223
• collapse short paras Addressed in More cleanup to episode 12 #224
• Move "Explore The Data’s Source" to answer for challenge Addressed in More cleanup to episode 12 #224
• Explain melt() function call.
• Explain facet.wrap. First time learners have seen?

Episode 13

• "Make sure you have" these libraries loaded Addressed in Cleanup to last two episodes #225
• Add loading in data step to start of file, echo = FALSE Addressed in Cleanup to last two episodes #225
• Add "we will continue working with . . . [ files ] " at start of file - stylistic consistency
• Remove first plot code and data tip (redundant with previous episode). Addressed in Cleanup to last two episodes #225
• Change title to reflect content - this is really about customization and creating publication quality graphic. Somewhat redundant with second half of Episode 7 and Episode 8. Addressed in Cleanup to last two episodes #225

Episode 14

• "Make sure you have" these libraries loaded
• Add loading in data step to start of file, echo = FALSE Addressed in Cleanup to last two episodes #225
• Add "we will continue working with . . . [ files ] " at start of file - stylistic consistency
• data.frame to "dataframe" Addressed in Cleanup to last two episodes #225
• change calculating mean NDVI to use pipes
• remove nested function alternative for NDVI calc Addressed in Cleanup to last two episodes #225
• remove the "# view only the value in row 1, column 1 of the data frame" cell Addressed in Cleanup to last two episodes #225
• move "# note the use of the vertical bar character ( | ) is equivalent to "or"." to callout Addressed in Cleanup to last two episodes #225
• remove single line alternative code chunk Addressed in Cleanup to last two episodes #225
• [ ] reverse order of variable assignment so that "->" always points the same direction (cognitive load)
• Move "For more information on date-time classes" to callout Addressed in Cleanup to last two episodes #225
• Instructions for "NDVI for the San Joaquin Experimental Range" are not clear. Does it want a comparative plot of the two sites? Not shown in answers given. From rest of episode, it looks like it just wants learners to calculate the averages and create the dataframe as done for the other site. Clarify instructions. Addressed in Cleanup to last two episodes #225
• Remove "We will use the ggplot() function within the ggplot2 package for this plot. If you are unfamiliar with ggplot() or would like to learn more about plotting in ggplot() see the episode on Plotting Time Series with ggplot in R." Addressed in Cleanup to last two episodes #225
• Explain for loop? Or use faceting instead? Addressed in Cleanup to last two episodes #225
• break up code chunks and move comments to narrative Addressed in Cleanup to last two episodes #225
• write.csv can be set to exclude row names, don't need extra lines of code

[jsta]

Too much typing can be a real struggle for workshops good idea to minimize it.