I have data from the Census Bureau's LEHD files that record number of jobs within blocks groups. I want to convert the number of data in the LEHD into a feature that I can use to calculate block density of the density of jobs in the block groups. How do I do that? I have read something about using the Float to Raster tool but I am not sure how to use it.

original data

joined data

  • Can you provide a screenshot of your data, or better yet, a link to a sample data set? Jul 15, 2014 at 18:38
  • I have attached the data. Basically the highlighted column has only "1" which represents a point so i need to get Kernel density by block. The number type in "float"
    – Dr.Apell
    Jul 15, 2014 at 18:49
  • I'm unclear on what you're trying to do. You want an attribute for each block that is total jobs/area? And right now you have duplicate polygons for each block, one for every job there? Your screenshot indicates a join has been done, but not what kind or how. Screenshots of your original data might be helpful. I'm not seeing a reason to convert anything to points, unless you're looking for one point with a total job count per block to then run a KD on.
    – Chris W
    Jul 15, 2014 at 20:24
  • What does the data look like on a map?
    – Erica
    Jul 15, 2014 at 20:31
  • Ok, I added the original data I hope I did it right since I am new to the site. The smaller table is the census data in column "SOOO" the larger table is the att table of block shape files. The goal is to create a buffer of 1 mile and calculate the density of 1's within the radius. I hope this makes it clear.
    – Dr.Apell
    Jul 15, 2014 at 20:50

3 Answers 3


I've worked with LEHD data before, including in a GIS context so I can probably clear up some of the confusion. First off, S000 does not represent a point, it represents an attribute. Secondly, I can tell that user26056 is working with Origin-Destination data because the GEOID's in the w_geocode are repeating and the S000 usually shows a bunch of 1's for almost all records. Third, S000 in the context of the Origin-Destination dataset represents the number of jobs held by people who live in the 2010 Census Block defined by h_geocode and also work in the 2010 Census Block defined by w_geocode. Sometimes S000 equals more than just 1, although this probably happens in densely populated areas.

If you are interested, I made a tutorial back in Grad School along with accompanying example data that shows how to create heat maps using Kernel Density in ArcGIS. I hope it helps your situation.

  • Great ! This answers my questions. Thanks to you and all those who tried to help. I know my explanations were probably not helping buy luckily this member was familiar with the data. Thank you all!!
    – Dr.Apell
    Jul 17, 2014 at 3:29

You do not explain what kind of data you have, but if it is point data you should be able to use QGIS to do point on polygon analysis. Here is a link to a article about this: http://maps.cga.harvard.edu/qgis/wkshop/pt_in_pgn.php

BTW - Census block data are stored as vector polygons, so you should not need to transform your data to raster

  • The data in the columns are numbers. As i said above a "1" represents a point. So what i am trying to do is convert the number one into a format that i can calculate density
    – Dr.Apell
    Jul 15, 2014 at 18:41

A shapefile consists of 3 (ish) parts: The .shp, which contains the geometry, the .dbf, which contains the datatable, and the .shx, which relates them to each other. To get density, you will need a shapefile with density, which is a polygon. LEHD data is provided in points. However, there is a field on the LEHD data has a field that can be used as a unique identifier to match it to census polygon shapefiles for 2010. It may be called FIPS or GEOID or something else (I forget). Download the polygon shapefiles from Census TIGER or NHGIS. Join the DBF from the LEHD data to that file. Add a field to the resulting file, calculate the area of each polygon. Divide the number of jobs by area to get density.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.