africanmathsinitiative / R-Instat-Help

The latest uncompiled (.hnd) and compiled (.cmd) versions of the help file for the statistics software package R-Instat
https://chuffed.org/project/africandatainitiative
GNU General Public License v3.0
2 stars 5 forks source link

Organising data for analysis #5

Open rdstern opened 7 years ago

rdstern commented 7 years ago

This is the possible video mentioned in the previous message #4 as number 3 there. I now have some suitable files from Cathy. This video is also one where we discuss the use of a spreadsheet and a statistics package for data analysis - which of course includes manipulation.

We have a series of examples, including one where we show how well you can organise the data using a spreadsheet - hence this can remain an option quality.xlsx

This file (called quality) shows a hand-written form and then the form entered into Excel. Another sheet shows data with a graph alongside. We would put the graph in the output window instead. Interestingly it is a graph of a boxplot, which needs an Excel addin. If you are going to get clever with Excel, by including a special addin, then perhaps an alternative is to consider a statistics package instead - or as well.
Then it has some examples of data in Excel that are ready for importing into a statistics package.

mauna_DataTranspose_working.zip

This is a zip file of an Excel workbook with a macro (called transpose) that actually stacks, rather than transposes the data. (Maybe we could rename it.) Excel doesn't have a stack feature, which all statistics packages have. However you can use the command language behind Excel to extend its capabilities. Useful, but that is the same sort of expertise that would also mean that you could use R, rather than R-Instat. Good example for discussion. If not you will be doing a lot of copy-paste, and could easily make a mistake.

Demo session 2.xlsx Excel_Figsdata.xlsx

These show a nice example which shows the use of comments and data entry into Excel - with checking, etc. It also shows metadata at different levels and sets the question of what will be allowed in R-Instat.

These examples form the basis for the first half of a video. We will give the results in Excel, without any instructions on how to do it. We are not teaching Excel here. However how to do it is in one for the good-practice guides and in our book - and we have permission to include our book free as pdf with the software.

A final example I would like to show is some climatic data (possibly from Ghana) this is a common format, where the columns are for each day of the month.

Before finalising the structure of this video I want to be clear of what features we plan to offer in R-Instat that correspond to current facilities in Excel, for example being able to add a comment to a cell. (Genstat has this feature also.)

The following comment will outline a possible structure of the video, but we need to be clear on proposed features in R-Instat first.