Overiew

In this Application I try to Plot the trend of AQI in Pune. And plot using Plotly and Base plotting system.

AQI

AQI transforms complex air quality data of various pollutants into a single index value which makes it easy to understand. Extent of air pollution and its severity can be mapped by calculating AQI for the respective air pollutants. Increase in AQI indicates increased air pollution and increased chances of threats associated with pollutants to human health.

Diffrent Level of AQI

(0-50)- good.
(51-100)- satisfactory, might cause little discomfort while breathing.
(101-200)- moderate pollution, might cause some lung diseases like asthma.
(201-300)- bad, will cause even more discomfort may also develop heart problems.
(301-400)- very bad, exposure to this air will cause severe damage and may need oxygen supply.
(401-500)- hazardous, this will create worst effect on bodies long time exposure will reduce life spam.

Link

Cleaning data

Raw Data

raw <- read.csv("Data/PNQ_AQI.csv")
clean <- read.csv("Data/Clean_AQI.csv")
head(raw)

                 Date SO2.Âµg.m3 Nox.Âµg.m3 RSPM.Âµg.m3 SPM CO2.Âµg.m3 AQI
1 2009-01-01 00:00:00         15         53         179  NA         NA 153
2 2009-02-01 00:00:00         15         48         156  NA         NA 137
3 2009-03-01 00:00:00         13         51         164  NA         NA 143
4 2009-04-01 00:00:00          8         37         135  NA         NA 123
5 2009-07-01 00:00:00         13         36         140  NA         NA 127
6 2009-08-01 00:00:00         10         30         135  NA         NA 123
  Location
1  MPCB-KR
2  MPCB-KR
3  MPCB-KR
4  MPCB-KR
5  MPCB-KR
6  MPCB-KR

There are 8 Columns and 7844 observations from 11 different locations. This data is recorded between year 2009 to 2019 Problems in this data is there are two formats of date and it need to be cleaned . I cleaned that format also selected only AQI, Date, and Location columns for ploting as we don’t need other columns we drop it. You can whole cleaning code in github repo. While ploting I also changed Location variable in usable format using make.names function

Cleaned Data

head(clean)

  X       Date AQI Location
1 1 2009-01-01 153  MPCB-KR
2 2 2009-01-02 137  MPCB-KR
3 3 2009-01-03 143  MPCB-KR
4 4 2009-01-04 123  MPCB-KR
5 5 2009-01-07 127  MPCB-KR
6 6 2009-01-08 123  MPCB-KR

Time series Plot

You can see there is sinusoidal trend in th data. Server in the app take input of radiobutton and fillter data for that sensor from whole data table. And then plot appropriately. Next we can look at decompose of this time series.

# filteredData is data filltered from user input of sensor of choice with Radio button
g <- plot_ly(
            filteredData(),
            x = ~ Date,
            y = ~ AQI,
            type = 'scatter',
            mode = 'lines',
            color = "#D9077E"
            )

Pune AQI (2009 - 2019)