Closed nniiicc closed 1 year ago
Hi professor Weber @nniiicc, I have scraped the Gubernatorial races, but I have just realized that the original scrape only got the Twitter handles not the Facebook, Twitter, or Youtube links. I was wondering which kind of information would you prefer to have.
Hi - it would be great to copy all of them - but if we just have twitter handles that is perfectly ok. (Don't spend more than 1 hour trying to go back and get the others)
Oh - also - Can you check in the scraper code for the Gov races?
Oh - also - Can you check in the scraper code for the Gov races?
For sure, I'll push it now.
Hi professor Weber @nniiicc . Since Municipal elections contain both information for mayoral and city elections, do you want information for these two elections separately on two data frames or together in one?
Good question! Either option is fine. Separate data frames might be easier for analysis, but you could also just add a variable about the race into the data frame (like below)
Candidate Name, City, Race, Balltopedia link, twitter link....
Bruce Harrell, Seattle, Mayor, HTTP://somelink.com,@BigBruce
Good question! Either option is fine. Separate data frames might be easier for analysis, but you could also just add a variable about the race into the data frame (like below)
Candidate Name, City, Race, Balltopedia link, twitter link....
Bruce Harrell, Seattle, Mayor, HTTP://somelink.com,@BigBruce
Got it! I will try to put them in separate data frames. I also have another question @nniiicc . I figured that there are many different races under the municipal and local elections, including county elections, county attorney, county clerk... And these races are different based on the county. I was wondering if we want information from all these races or only the Mayoral and City elections.
We want data from all races... This is why it might be easier to just add a variable about which race the candidate is in ... Does that make sense?
We want data from all races... This is why it might be easier to just add a variable about which race the candidate is in ... Does that make sense?
Got it! Since different link has different layout, so I'm thinking maybe it's actually easier to treat them separately. My current method would be to have separate data frames for mayoral and city elections while having all the races under municipal elections together in a data frame. Does that sound ok?
Sounds great!
Hi professor Weber @nniiicc , I have a question about this type of layout. https://ballotpedia.org/City_elections_in_Anchorage,_Alaska_(2022) I was wondering if we need the "office" information in the Anchorage Assembly section. I have figured out how to get the race name and the candidates under each race, but not which seat they are competing for... If we need that information I would need some help.
Just the race name is fine - thanks for asking (that's a weird election set-up)
Yeah, it's confusing. I just realized the Anchorage Assembly, city attorney, judicial offices, etc. are all part of the city council. And each city would have different offices for election. To make them consistent for analysis, should I just put them as city council? @nniiicc
hmmm - I'm not sure I follow. I think Anchorage has a number of races (e.g. assembly, city attorney, etc) ... If you scrape them all and put them all in the same dataframe its perfectly fine - just be sure that candidates are labeled by race
Got it! I have another question about the primary and general elections. For city council elections the only time the general election takes place is when 4 people are selected as primary and only three can win. Two of them will fight for one spot. In this case, should I also get the primary information?
Hmmm ... this is tricky. I would say that no we should not collect primary data. We can just skip those races where a general election didn't take place. Good catch!
Got it, and I have reuploaded the Mayoral result without primary information. The city election has multiple different styles, and I have got most of them covered just need to figure out how to ignore the primary cases now. The municipal election for the county is similar to the city one, so I think it will be quick after I got the city covered. One question about the municipal election for the county, do we want the district information as well? Like which district the candidate is running for.
Yes, if we can include district information that would be great.
The next process is to evolve the scraper to collect more data across more races. Below are data sources and races we want to collect the data from: