Data Skills Class

[art-of-eons]

If you look over the CDC website and the data they’ve compiled for US deaths and their listed causes you’ll notice a “weird” trend. Every-time a death certificate is issued, the cause of death must be listed or identified. Historically speaking, roughly 99% of death certificates clearly delineate a “cause” or reason for the death. The other 1% (roughly 500-600 deaths per week) is entered as a “mystery” death.

However, starting in 2020, the number of deaths in this “mystery” category has dramatically spiked.

A weird, “mystery” illness (that’s not being listed as Covid 19) is possibly killing up to 3000 Americans a week. And in many of the states that are pushing to reopen, this number seems to be skyrocketing.

Now if states like Florida, Texas, Georgia, etc. want us to believe that this massive uptick in unclassifiable deaths is somehow unrelated to Covid 19 - they still have to explain how hundreds of people mysteriously dying each week isn’t a public health crisis.

[dataskills-class]

If someone posts data but doesn’t name their source (the CDC isn’t a source, it’s a massive agency that maintains dozens of public data sets), that’s your first clue that they’re fudging things. If you actually look at the CDC mortality data, they suggest a lengthy citation:

“Centers for Disease Control and Prevention, National Center for Health Statistics. Underlying Cause of Death 1999-2019 on CDC WONDER Online Database, released in 2020. Data are from the Multiple Cause of Death Files, 1999-2019, as compiled from data provided by the 57 vital statistics jurisdictions through the Vital Statistics Cooperative Program. Accessed at http://wonder.cdc.gov/ucd-icd10.html on Mar 17, 2021 11:33:00 PM”

CDC’s WONDER is where mortality data is tracked and as of today (March 2021) there is no public mortality data for 2020 available. When this was posted back in July, that data definitely wasn’t available. WONDER is available as monthly data, but not weekly. For weekly mortality data you need the Morbidity and Mortality Weekly Report (MMWR), but I’m not seeing “Not elsewhere classified” there.

It looks very official that they cite “R00-R99″ on that top chart, but if you actually look at WONDER you’ll see that the range includes Asphyxia (R09), Anorexia (R63) and SIDS (R95). There are, as you might expect a hundred different causes of death between R00 and R99.

R99 itself is the bucket for “Other ill-defined and unspecified causes of mortality”. Which is not the same as “Not elsewhere classified” which isn’t a thing. “G98 (Other disorders of nervous system, not elsewhere classified)“ is a thing as is “I51.1 (Rupture of chordae tendineae, not elsewhere classified)”. There are a bunch. R93.0 (Abnormal findings on diagnostic imaging of skull and head, not elsewhere classified) or W24 (Contact with lifting and transmission devices, not elsewhere classified) are things.  Reading the list of possible causes of death is definitely disturbing. (”Cutaneous abscess, furuncle and carbuncle of buttock” was responsible for 38 deaths in 2019)

[dataskills-class]

Earned v. Viral

[analyticisms]

#FridayFunFact: the debate about whether teens are leaving Facebook for greener social media pastures still rage son.

[kenyatta]

lol misunderstanding g+ numbers.

[otter-ottter]

What I find hilarious here is … Snapchat? ask.fm? Kik?  I’m impressed that there are more teens on LinkedIn than Reddit, but … get with the times.

[kenyatta]

Yeah, I haven’t downloaded the report yet but Snapchat seems to be in there.

Not seeing Snapchat in the top 10 could be a US vs Global perception thing on our parts. Same with the appearance of LinkedIn which Euro/Latino/Aussie adults treat a lot more seriously so maybe baby thems feel like they have to create their own profiles(?)

I can’t remember if GWI reports exclude China but I would expect to see a different top 10 list here if it did.

[dataskills-class]

cc @advocacyjourno

[themomsmustbecrazy]

I am concerned about the narrative Serious is coming home with after lockdown drills at school. Lockdown drills come out of a very disturbing narrative, so I guess that’s not surprising, but I still wish I could find a way to push back.

I do understand that it lockdowns are part of required...

[dataskills-class]

cc cfurlong2006

[dataskills-class]

DUI's by state

The states with the most DUI arrests per 100,000 residents are, more often than not, sparsely populated.  Wyoming, Alaska, and North Dakota are not much of a surprise, but Wisconsin's penchant for drinking and driving is something I've heard of while never seeing empirical evidence.  

Scott Walker's policies are driving his constituents to reckless behavior and he needs to be stopped.  

[dataskills-class]

Migration Waves in New York City

We want to gather historic Census data, from 1850 to this day, from  8-10 different migrant populations, and try to answer some things:

- Why did they come to NYC? (War, diseases, economic opportunity)

- Where did they settle in the city and why (it would be interesting to show, for instance, when Little Italy became an Italian neighborhood)

- Did they move after? And why?

The big driver of the story would be something like an animated GIF of a NYC map that would show shapes of migrant populations, color-coded for the country of origin. To accompany that, a timeline showing some milestones of migration movements. And, lastly, some simpler line graphs with the different waves for different groups (tentatively: Italians, Irish, Russians, Chinese, Mexicans, Germans, Dominican Repubic, Jamaica).

Besides making a visually interesting (and shareable) GIF, the story will be interesting to show and verify what neighborhoods are being gentrified.

DATA SETS:

http://www.city-data.com/states/New-York-Migration.html

(City Data on waves of migration to NYC)

http://www.gothamgazette.com/index.php/city/2738-immigrant-city

-By Sean Devlin and Pedro Burgos

[dataskills-class]

[dataskills-class]

I found an intersting dataset from the Information technology and innovation foundation about the economic evolution and developpment around the United states for the IT sector. I download the...

[dataskills-class]

The 2014 State New Economy Index is comprehensive report on the American economy and in one section examines the energy production/consumption of each state: The Clean Energy Economy.

The color gradient shows how much each state produces in clean energy as a total percentage of the state's consumption: dark red states (ME, WA, OR, MO) are the leaders with the highest proportion of clean energy while legacy oil states such as TX are light yellow.   

[dataskills-class]

Mapping Mental Health Treatment Centers in the US

One of my goals with my community, African-Americans suffering from depression, is to not only build a safe forum for discussing mental health, but to also provide resources for seeking treatment. After a couple of hours (and yes, I do mean hours) I finally found some useful, relevant, mappable data.

While on the Substance Abuse & Mental Health Services Administration’s  (SAMHSA’s) website, probing for a useful dataset file, I discovered a treatment facility locator. The locator is updated weekly and lists, with several exceptions, most public, private for-profit, and non-profit State-licensed or nationally accredited mental health facilities around the nation. Another plus? The locator allows you to download a real-time .csv file of the records. And even better, you’re able to filter the facility locator data, which is perfect when searching for specific types of facilities (e.g. a Public/government sponsored facility that accepts Medicaid) or services for a specific demographic (e.g. treatment for adolescents or LGBTQ patients). A useful resource for my community? You bet.

I will use the data to generate a map of shapes and will then merge it with the US State shapefile in CardoBD, listing facilities per state. I'll most likely be using the common field “state.” If time permits, I'd like to also try generating a NY-based map using a NYS zip code shapefile.

- Adriele Parker

[dataskills-class]

Mapping America's Clean Energy Economy

Using the excel data set from the 2014 State New Economy Index I'm mapping the clean energy economy in the U.S., which is calculated as a state's clean energy consumption as an overall proportion of energy consumption. 

I'm focusing my community project on the environmental technology startups in New York and this will help me to visualize how the entire country is working, or not, towards shifting state economies toward renewable energy.   

[emilygoldblum]

[dataskills-class]

Using the limited data that’s available on same-sex couples in the U.S. I’d love to map out where the highest number of same-sex couples live. I believe I’ll have to use a similar tactic to how we…

http://cdb.io/1wgZD7E

Update** After moving around the data in an excel, I found the best solution were to take the state column, number of same sex households and narrow the percentage to households that have same sex spouses vs just those that have same-sex persons living in that household. I was able to take the geometric portion from the map we used last week (US State Shapefile) and merge it with the CSV I formatted. The results showed a map that presented the percentages of same-sex spouses of the United States. This is according to data compiled in 2013.

[dataskills-class]

Mapping the Movement's Struggle in Missouri

My project is focusing on how Black trans, LGB and women fit into the Black Twitter activism and millennial community. And although I'm compiling data on these social issues, I'm also compiling on the Black Lives Movement (sparked by Twitter activism) and the issues it's currently focusing on: Ferguson, racial profiling, police brutality and endangerment of Black male lives.

The data I'm mapping will show the location in major Missouri cities/counties where Black folks were stopped by police compared to white folks. I'm going to use a Missouri-shaped map to represent the geographical location. Also, I'm going to use data shaping with the colors of blue (Black folks) and red (White folks) to show the percentage of Black folks stopped compared their White counterparts. 

Another option will be to show it through a radiant of color -- comparing the Black population percentage simply to the number of Black folks stopped.

Or is there a way to map both together -- Black versus White stops by police and the population of Black and White folks in major Missouri counties?

Hmm...

Actually, I realized it was impossibly hard to map racial profiling as I needed location to prevalent in the data. 

After a tiresome search, I decided to map general data about Internet usage across the United States. The data shows the amount of Internet usage anywhere within each state. Internet Usage was most prevalent in California and Texas.

Here's the map: http://cdb.io/1zxYpkh

[dataskills-class]

Pitch: this map shows that we knew in 2009 that the water crisis was coming in Brazil

Brazilian government is blaming climate change and the severe drought for the current water shortage that people from all over the country – particularly in the rich Sao Paulo state – are experiencing. But some environmentalists have pointed out that we know that we needed a better water infrastructure for years. And some parts of the state needed those fixes more badly than others. 

Luckily, Brazilian Federal Water Agency (ANA) released a study in 2009 showing their assessment in each municipality. In Sao Paulo state, most of them were marked as "satisfactory water supply", but big cities like the capital, Sao Paulo, received a "requires extension of the current system" mark, and some even "requires a new water source".

My idea is to plot in a map this information, color-coding where, in 2009, we knew that the situation was going to get bad. I could insert more data-points later (like how was the situation 5 years after, or how much money was invested), but this single map would be sufficient to point that we could have acted way sooner to prevent the calamity.

I found exactly the spreadsheets that I needed here (a list of municipalities in row B and in row W what was needed to be done in 2009). Now I need a detailed map of Sao Paulo state in Brazil (that I think that I can find here), so I can use the municipalities name with the geographic borders.

[dataskills-class]

[dataskills-class]

I’m going to map one of the data points from our community profile pitch. I will look at the number of undocumented immigrants eligible for deportation relief under the Obama administration’s...

Some answers:

+ Does the quantile method illustrate differences or exaggerate them? 

+ We can use this map as our number formatting example in class next week (in other words, yes).

+ The last two, we'll go over in class.

[dataskills-class]

[dataskills-class]

I recently had a meeting with Kim Wiley Schwartz, Assistant Commissioner for Safety Education and Outreach for the NYC Department of Transportation. The meeting was held to find out why DOT had closed...

I should be much more clear with everyone that I’d like you to actually link to the JSON file you’re using. But I happen to know the file you’re speaking of and I know it already includes lat/lon information. 

A great first map.

[dataskills-class]

Mapping American Ancestry in the Mid-Atlantic States

For my mapping project, I decided to not only work with the Irish community that is part of my greater project, but gain a scope of all the ancestries that comprises the states of the Mid-Atlantic.  I'd be using shapes for this undertaking.  Additionally my data comes from the ACS 2013 estimates, and I'll look at specific ethnicity groups in specific states by using Refine.  Additionally, I'd like to use this to figure out what percentages of groups from other countries make up the largest percentages of state populations.

-Sean Devlin

[dataskills-class]

[dataskills-class]

This might be the angriest blog I’ve ever written or not… But it seems as though the nature of this work isn’t hard, but doesn’t want me do it, if it makes sense. The simplest errors and tasks become...

Uff. That does not sound fun.

It does help to be systematic. Even starting to type out pleas for help is a good approach. Spell out your steps, at least to yourself, the way you might if you know your keys are here somewhere. Last night, I walked in the door, stepped in a pile of legos, hopped to the cou…Oh! my keys are on the couch. But you shouldn’t be spending hours fighting with software — ask for help if you’re stuck! Even consider posting something here.

And, if you had any “aha” moments that might help someone who thinks like you next year, I’d love to update / expand my troubleshooting suggestions!