Feeling threatened? Shoot them.

[merphie]

Too bad. Government agencies should be required to publish data as data.

On some of the stuff they give a data table, but it's not in a CVS format or anything. It's hard to get anything out of the government. They like to take information but not give it back

 

[CFLarsen]

Too bad. Government agencies should be required to publish data as data.

Why don't we just agree on this:

If you want to use data, you have to accept that they are to be examined. Ergo, you have to present it in a format that is suitable for examination.

If you refuse to provide a suitable format, you lose by default.

I suggest we can submit data in the following formats: CSV and XLS. Of course, once you got CSV, you got XSL.

Any objections?

(I opened a new thread for this here.)

 

[varwoche]

They offer an annual publication called "Crime in the United States." I'm not sure if the data you're looking for is there. The file is available as pdf, and the tables as xls.
http://www.fbi.gov/ucr/ucr.htm#cius

Bingo, XLS will do just fine. Thanks.

Here's the data available for 2003. This looks ripe for exploration.

Table 1 - Crime in the Unites States, by Volume and Rate, 1984 - 2003
Table 5 - by State, 2003
Table 6 - By Metropolitan Statistical Area, 2003
Table 8 - Offenses Known to Law Enforcement, by City 10,000 and over in Population, 2003
Table 8A - by City under 10,000 in Population, 2003
Table 10A - by Non metropolitan County under 25,000 in Population, 2003
Table11A - by Other Agency by State, 2003
Table 2.10 - Murder Victims by Age, by Weapon, 2003
Table 2.11 - Murder Circumstances, by Relationship, 2003
Table 2.12 - by Weapon, 2003
Table 2.13 - Murder Circumstances, 1999-2003
Table 2.14 - Murder Circumstances, by Victim Sex, 2003
Table 2.15 - Justifiable Homicide by Weapon, Law Enforcement, 1999-2003
Table 2.16 - by Weapon, Private Citizen,1 1999-2003
Table 2.2 - Murder by Month, Percent Distribution, 1999-2003
Table 2.3 - Murder Victims by Race and Sex, 2003
Table 2.4 - by Age, Sex, and Race, 2003
Table 2.5 - Murder Offenders, by Age, Sex, and Race, 2003
Table2.6 - Murder Victim/Offender Relationship, by Age, 2003
Table 2.7 - by Race and Sex, 2003
Table 2.8 - Murder, Types of Weapons Used, Percent Distribution by Region, 2003
Table 2.9 - Murder Victims, by Weapon, 1999-2003
Table 69, Arrests, by State, 2003

http://www.fbi.gov/ucr/03cius.htm

 

[varwoche]

Why don't we just agree on this:

If you want to use data, you have to accept that they are to be examined. Ergo, you have to present it in a format that is suitable for examination.

If you refuse to provide a suitable format, you lose by default.

I suggest we can submit data in the following formats: CSV and XLS. Of course, once you got CSV, you got XSL.

Any objections?

(I opened a new thread for this here.)

I have something more useful (but more involved) in mind:

1) create a schema (rdbms) to contain all of the relevant data
2) if there are other rdbms types amongst us, we can discuss the schema before I implement it
3) import the XLS files into rdbms (The formats vary from year to year so this will be the biggest hassle. It needs to be done though -- if we don't consider how the data is trending then it will be a waste of time imo.)
4) I'll generate any analysis that anyone requests if it's doable/feasible, including charts
5) For anyone who wants to do their own analysis, I'll generate extracts in CSV, XLS or whatever -- all I need is a place to upload.

 

[T'ai Chi]

Yeah, I can type in a city or anything else. Problem is, if I type in "Los Angeles" AND "permit", I get 0 hits.

39 = 0? Must be the "new math".

You have to show your data, because you have a history of cheating.

No, he really doesn't. You do have a history of bullying, however.

 

[CFLarsen]

39 = 0? Must be the "new math".

Feel free to show me the search criteria.

No, he really doesn't. You do have a history of bullying, however.

That is not the site that shanek was talking about. We are talking about the packing.org site.

Please keep up here. Or, at least, make an effort.

 

[shanek]

I'll explain, but first I need to see if the the raw data is available. The department of labor provides vast quantities of downloadable raw data. Like this.

The UCR is available in several different Excel files. Sorry I can't link to them, but it looks as if the pages are dynamic, and of course dynamic pages don't bookmark.

Has anyone noticed if the gun stats are similarly downloadable?

Not that I've seen. I've had to pour through many different pages trying to find them.

 

[Flex]

Just a thought,

It may be appropriate to clearly, and unabiguously, state the question we are attempting to answer.

-Flex

 

[CFLarsen]

shanek,

Any chance of you showing your data?

 

[T'ai Chi]

shanek,

Any chance of you showing your data?

Why don't you? You could use it as an opportunity to present the usefulness of the Excel format you were theorizing about.

Perfect opportunity to show us your stuff.

 

[CFLarsen]

Why don't you? You could use it as an opportunity to present the usefulness of the Excel format you were theorizing about.

Perfect opportunity to show us your stuff.

Knock yourself out.

Now, will you also ask shanek for his data?

 

[varwoche]

Bingo, XLS will do just fine. Thanks.

Here's the data available for 2003. This looks ripe for exploration.

http://www.fbi.gov/ucr/03cius.htm

Bummer... The data is pre sliced and diced and they don't publish the raw data. This means for instance we can look at stats by state, and we can look at stats by murder weapon, but we can't look at stats by murder weapon by state.

 

[CFLarsen]

Bummer... The data is pre sliced and diced and they don't publish the raw data. This means for instance we can look at stats by state, and we can look at stats by murder weapon, but we can't look at stats by murder weapon by state.

It's a start.

 

[crimresearch]

Why don't you? You could use it as an opportunity to present the usefulness of the Excel format you were theorizing about.

Perfect opportunity to show us your stuff.

I too would like to see Claus answer some of the standing questions using that data.

Perhaps he would like to start by telling us how he plans to control for multicollinearity, autocorrelation, skedasticity, etc. in drawing his causal inferences.

 

[shanek]

Bummer... The data is pre sliced and diced and they don't publish the raw data. This means for instance we can look at stats by state, and we can look at stats by murder weapon, but we can't look at stats by murder weapon by state.

No, but you can see the data as I've broken it down: by state, and by type of crime (murder, robbery, etc.). There's also an excel file with the same breakdown, for all cities with a population greater than 10,000. That's what I used to make my most recent posts.

I, for one, would like to see the number of robberies that occured when someone was present as opposed to when there was no one around.

 

[CFLarsen]

I too would like to see Claus answer some of the standing questions using that data.

Perhaps he would like to start by telling us how he plans to control for multicollinearity, autocorrelation, skedasticity, etc. in drawing his causal inferences.

Perhaps if you list them, and explain why I need to control for "multicollinearity, autocorrelation, skedasticity, etc", when you don't demand this of others.

And don't forget to expand on that "etc".

 

[CFLarsen]

No, but you can see the data as I've broken it down: by state, and by type of crime (murder, robbery, etc.). There's also an excel file with the same breakdown, for all cities with a population greater than 10,000. That's what I used to make my most recent posts.

I, for one, would like to see the number of robberies that occured when someone was present as opposed to when there was no one around.

I, for one, would still like to see where, precisely, you have found the data about cities and their gun legislation.

We need to check your data, because you have a history of cheating.

 

[T'ai Chi]

Knock yourself out.

Now, will you also ask shanek for his data?

I wasn't aware the FBI data was yours.

Your imagined connections must run everywhere.

 

[T'ai Chi]

I too would like to see Claus answer some of the standing questions using that data.

Perhaps he would like to start by telling us how he plans to control for multicollinearity, autocorrelation, skedasticity, etc. in drawing his causal inferences.

He'll never answer anything specific. Just watch and enjoy the show.

 

[T'ai Chi]

I, for one, would still like to see where, precisely, you have found the data about cities and their gun legislation.

We need to check your data, because you have a history of cheating.

You already said that.

No, he doesn't, as already said, but you do have a history of bullying, as already said.

So you post a link to an FBI site that has data. SO WHAT? So what are you going to do about it? Are we supposed to be impressed with a link to a site?

Whatever..