Read this document first to get started. This applies for fixed-width data procured using the old method. I have made instructions on how to use the data in the new format(s) and modify the scripts here to work with that. But read this first before going there, as that document assumes that you did.
-
Downloading and installing an Integrated Development Environment (IDE) for R such as RStudio is highly recommended.
-
Obtain the Data: Get the unit level raw data along with the documentation for this survey. Some links to download this data are provided in the README file, but the process might be different at the time of your viewing. Reach out if you have queries.
-
Prepare lists: In the future I may document how to use NSSO's documentation to prepare some lists provided in this repo to help with estimation. For now the lists I have provided can suffice. However, I strognly suggest the beginner to go through NSSO's documentation to prepare these lists themselves.
Here is a quick and basic introduction to R (although without any license information). Here are some cheat sheets from RStudio. Here are some directions on using RStudio and GitHub together if you are into that sort of thing.
Decide on a working directory (a folder/directory in which you will do the work). As an example I will use /home/user/Desktop/SAS2019 as the working directory. The scripts provided here assume the following folder structure in the working directory:
/home/user/Desktop/SAS2019/Raw Data: All the raw data are stored here, including all levels from both visits./home/user/Desktop/SAS2019/Output: This can be an empty folder to begin with. It will get populated with all the calculated data frames in the form of .csv and .RData files as we execute the scripts.
The R scripts in this repository contain the code and comments for reading, estimating, and conducting basic analysis of the unit-level data from Schedule 33.1 of the 77th round by NSSO.
The analysis can be conducted in the following order:
| File Name | Description |
|---|---|
| 01_All_Basic_HH.R | Creates a dataset with some basic information about all the households, and identifies agricultural households that are common across two visits. |
| 02_Land_Size_Class_Combined.R | Adds land size class to the data. |
| 03_Crop_Income.R | Calculates the crop income using the paid-out costs approach. |
| 04_Animal_Income.R | Calculates animal income using the paid-out costs approach. |
| 05_NonFarm_Business_Income.R | Calculates income from non-farm businesses. |
| 06_Other_Income.R | Calculates other incomes such as Wages, Rents and Pensions. |
| 07_HH_Income_of_AH.R | Combines crop income, animal income, wage, rent, and non-farm business income to calculate household incomes. |
| 02A_Land_Size_Class_Visit2.R | Optional script to calculate land size class from Visit 2. Land size class from visit 1 alone is calculated otherwise. If you are using this script, please uncomment the relevant code block in 06_HH_Income_of_AH.R to add that to the final data frame. |
These are the supporting files prepared based on the documentation, for use in the scripts.
| File Name | Description |
|---|---|
| List_Level_Codes.xlsx | Contains the widths of variables for data extraction, specified by NSS in the documentation. This is the only list that is necessary. The rest are optional. |
| List_State.xlsx | Contains a list of States against codes. |
| List_Social_Group.xlsx | Contains a list of social groups against codes. |
| List_HH_Classification.xlsx | Contains a list of household classifications against codes. |
| List_Religion.xlsx | Contains a list of religions against codes. |
| List_Crop_Code.xlsx | Contains a list of crops against codes (it has two lists - one for MSP Crops alone, and one for all crops in the survey.) This list is not used here, but will be of use to analyse cropwise data. |
There are two ways of going about running these scripts I will describe here. The first one is easier if you have some basic familiarity with git, and the other is for if you do not want to be learning about git.
- Clone the git repo to your desired location. (You can do this within RStudio as well.)
- Create the folders
Raw DataandOutputin the location. These folders are already added to .gitignore - Place all the fixed width files of raw data in the
Raw Datafolder. - Open the scripts in the IDE of your choice such as RStudio
- Read through the scripts.
- Install relevant libraries if required. (This can be done by running the code
install.packages("packagename")in the R Console) - Run them in the order of naming.
- Decide on the working directory. In this example I will use
/home/user/Desktop/SAS2019(Note: If you are in a Windows machine, you will need to reverse the slashes from, for example,D:\Statistics\SAS2019toD:/Statistics/SAS2019) - Bring all the scripts and the lists in this repository and place them on the working directory (You can just right-click and save the links from the main page of this repository)
- Create the folders
/home/user/Desktop/SAS2019/Raw Data/and/home/user/Desktop/SAS2019/Output - Bring all the fixed width files of raw data and place them in
/home/user/Desktop/SAS2019/Raw Data/ - Open the first script using an IDE such as RStudio.
- Read through the script.
- Install relevant libraries if required. (This can be done by running the code
install.packages("packagename")in the R Console) - Change the line
setwd(".")to the path to your working directory. - Run the script.
- Repeat the steps 4 to 8 for all the scripts in order.
You are done. The outputs will be generated in the Output folder.
You can of course change the folder structure to your liking so long as you make appropriate changes to the scripts as well.
If you have data in other formats, read these instructions for some help on how to modify these scripts to suit that.
Thanks for reading. Please reach out if the instructions are not clear enough. Feel free to open an issue if you come across any.