mirkomantovani / chicago-realtime-aq Goto Github PK

Sometimes the plot shows 12 hours back data, indicate to the user that current data is not available, most recent data is being shown

Warning: Error in map: could not find function "map"
175: function_list[[i]]
174: freduce
173: _fseq
172: eval
171: eval
169: %>%
168: get_and_preprocess_observations_24h_ds [/Users/mirkomantovani/UIC/Visualization and Visual Analytics/Chicago-RealTime-AQ/app.R#1756]
167: renderPlot [/Users/mirkomantovani/UIC/Visualization and Visual Analytics/Chicago-RealTime-AQ/app.R#1850]
165: func
125: drawPlot
111: reactive:plotObj
95: drawReactive
82: origRenderFunc
81: output$graphical_data_ds
1: shiny::runApp

Add another dataset that helps to augment and explain the air quality data in Chicago, that can also be queried live through R.
Ideas?

• Traffic data -> helps explain AQ through pollution created by cars

Created 3 options of map background

When everybody has finished their work, create a branch for normal screens and adjust all parameters and deploy on shinyapps.io.

This happens when clicking on light blue node and then another light blue node, it might be similar to issue #10 that I solved putting some other constraints like
retrieved_measures <- unique(df$measure)

if ("so2" %in% c(input$measures1,input$measures2) && "so2" %in% retrieved_measures){

and so on..

Error:
Warning: Error in : Aesthetics must be either length 1 or the same as the data (43): y
182: stop
181: check_aesthetics
180: f
179: l$compute_aesthetics
178: f
177: by_layer
176: ggplot_build.ggplot
174: print.ggplot
166: func
164: f
163: Reduce
154: do
153: hybrid_chain
125: drawPlot
111: reactive:plotObj
95: drawReactive
82: origRenderFunc
81: output$graphical_data_last_ds
1: runApp

Since we have to show the table with true/false and it takes up a lot of space but is not necessary to have it always there, we could put a switch on the sidebar to hide and show the table (this could be done via js easily as I was doing it with the project 2).

The main panel with data will probably have to be even bigger to show better the plots (bigger width, so we can't have 2 big panels always showing, no space for the map itself.

Check if the node has valid attributes and data before extracting data from it.

Warning: Error in <-: 'names' attribute [1] must be the same length as the vector [0]
168: get_and_preprocess_observations [D:\cs 424\Chicago-RealTime-AQ/app.R#639]
167: renderPlot [D:\cs 424\Chicago-RealTime-AQ/app.R#1455]

The OpenAQ nodes preprocessing needs to create a dataframe with the pm2.5 feature name as "pm2.5" and not "pm25"

We have to come up with a working way of storing the latest node click (in map and table independently) in the same variable so that it can be used easily in the future.

• After clicking on green (OpenAQ nodes), the plot should be updated by querying the OpenAQ data
• Maybe we will use a different panel, it could be easier and the requirements could be okay with that (Talk with TA)

170: send_request [D:\cs 424\Chicago-RealTime-AQ/app.R#422]
169: ls.observations [D:\cs 424\Chicago-RealTime-AQ/app.R#500]
168: get_and_preprocess_observations [D:\cs 424\Chicago-RealTime-AQ/app.R#636]
167: renderPlot [D:\cs 424\Chicago-RealTime-AQ/app.R#1455]

I don't know how it's possible but Dark Sky with my key has already done 1000 requests today and yesterday, how many different requests happen when clicking on a single node?
The error presented in the console is:

Warning: Error in get_forecast_for: Forbidden (HTTP 403).
181:

We might want to consider using different keys and switching randomly between them

As explained in week 3 of the course, it would be visually better to show a table in a form like:

Follow this guide to create a heatmap:
https://geodacenter.github.io/aot-workshop/Part2-AOT.html

Details:

• Color? Green
• No inactive nodes for OpenAQ
• Primary key is location -> changed name to vsn for simplicity to use "vsn" as a common node ID

Error for light blue marker on the map, when all measures data is not present.
Might need to check for data before subset the measure for plotting.

Conversion of all measures for all the plots to imperial from metric and viceversa based on switch input