Presentation on the cycle and tasks of data analysis
Ashu Guru
Author
03/13/2019
Added
89
Plays
Description
Presentation on the cycle and tasks of data analysis by Ashu Guru, Univ of Nebraska Raikes School (4 mins)
Searchable Transcript
Toggle between list and paragraph view.
- [00:00:00.560]Let's start looking into the analysis now.
- [00:00:04.160]The approach that we are going to take is we are going to work in a cycle.
- [00:00:08.640]So we enter, identify the task or tasks, then for each of those tasks we will
- [00:00:14.780]identify the function that the software provides or in this case the function
- [00:00:20.920]that R provides that will be able to perform that task for us.
- [00:00:26.280]Then in many cases, before you can actually apply a function
- [00:00:29.659]to a data, you have to prepare the data, you have to format it in a way that the
- [00:00:34.147]function can actually accept it.
- [00:00:36.400]Then we apply the function, we go through the next task,
- [00:00:40.350]we identify function, we prepare the data, apply function.
- [00:00:44.840]So this way we perform the analysis until we have completed all the tasks that we
- [00:00:50.050]identify and then we will exit this cycle.
- [00:00:55.800]Again.
- [00:00:56.320]Just a reiteration of the objective, which CSR indices and growth stages best
- [00:01:01.590]correlate with the yield and Oregon yield components.
- [00:01:05.760]Here is the look at the data from the sensor.
- [00:01:08.680]We have the wavelengths in column A, then we have readings for those
- [00:01:14.312]wavelengths as individual samples that are collected.
- [00:01:19.920]We also know that the instrument logs 2048 discrete wavelengths and then for
- [00:01:25.878]each wavelength it measures the reflectance.
- [00:01:32.640]These are the common spectral indices that we will be using.
- [00:01:36.760]The example that we will be working on will be using the water index WI which is
- [00:01:42.806]given by R 900 divided by R970.
- [00:01:46.520]The first task that we want to perform is that we have to filter and reduce the
- [00:01:51.602]data.
- [00:01:52.240]Let's look at what does it mean by filter and reduce.
- [00:01:56.520]So we have the raw data in this format in Excel.
- [00:02:00.480]We have to take this data and select the wavelengths that we are interested in.
- [00:02:06.400]As a scientist, we may be interested in averaging out
- [00:02:09.846]wavelengths which are close to 900.
- [00:02:12.400]So in this case, I could say that I am interested in the
- [00:02:16.133]band that I'm highlighting and find averages of these and consider it as the
- [00:02:21.176]900 wavelength for the calculation of WI index.
- [00:02:24.640]So we have filtered the data and then we have reduced it.
- [00:02:28.320]In this slide, we will get more specifics regarding
- [00:02:31.648]filter and reduce.
- [00:02:34.960]So from the reflectance data set of 2048 rows by N plots,
- [00:02:39.316]we want to pull out the wavelengths that would be used in any number of indices.
- [00:02:45.440]So for example, in the WI index, this would be the wavelength 900 and 970
- [00:02:51.959]nanometer.
- [00:02:52.840]Since wavelengths are not in one nanometer increments.
- [00:02:56.200]So the scientists would like to be able to systematically pull some numbers of
- [00:03:01.574]close wavelengths out to average together and then we want to save this as a new
- [00:03:07.085]data set or a new data file that is work in progress.
- [00:03:12.960]This is a review of the formula that is used for calculating water index which is
- [00:03:20.158]given by R 900 divided by R970.
- [00:03:23.920]This table shows us what we are expecting as the output after the task one is
- [00:03:29.734]completed.
- [00:03:30.480]So we should see that our data is similar to the format where in each row we have
- [00:03:37.277]averages of let's say 3 bands or five bands and 10 bands for 900 and 970
- [00:03:43.328]nanometers.
- [00:03:44.400]For each of the observation point, we have the averaged values that
- [00:03:49.875]correspond to the rows that are part of the average of each of the bands
- [00:03:55.753]wavelengths.
- [00:04:01.880]The task 2 is that we will then like to sort and transpose.
- [00:04:06.160]This is how we would like the data to be from the last task.
- [00:04:09.960]We want to convert it so that we just transpose rows into columns and columns
- [00:04:15.171]into rows.
- [00:04:16.080]This will help us in further analyzing.
- [00:04:18.840]And finally the task 3 is to calculate the correlation which is the output from
- [00:04:24.622]task 2 then correlated with the yield measures and that will give us the output
- [00:04:30.404]of the analysis that we are interested in.
The screen size you are trying to search captions on is too small!
You can always jump over to MediaHub and check it out there.
Log in to post comments
Embed
Copy the following code into your page
HTML
<div style="padding-top: 56.25%; overflow: hidden; position:relative; -webkit-box-flex: 1; flex-grow: 1;">
<iframe
style="bottom: 0; left: 0; position: absolute; right: 0; top: 0; border: 0; height: 100%; width: 100%;"
src="https://mediahub.unl.edu/media/10789?format=iframe&autoplay=0"
title="Video Player: Presentation on the cycle and tasks of data analysis"
allowfullscreen
></iframe>
</div>
Comments
0 Comments