Hello
So I helped to clean up the file PA -- 192xx.csv (link https://1fichier.com/?gqxdx82szi5lttdoehv2 last night and posted it back. The cleaned up version of that record, about 3MB, was saved here https://1fichier.com/?b0ijf36vi0wrrptp0fg2 It is a csv file and contains only the name, gender, date of birth, street address, city, and COUNTY, per sperg's request. Each column is labelled should be self-explanatory.
Got this file cleaned (PA192xx.csv -> PA192xx_cleaned.csv) https://1fichier.com/?b0ijf36vi0wrrptp0fg2
Please let me know if you can use it
Hello,
I have "PA192xx.csv" cleaned up, here is the result file PA192xx_cleaned.csv https://1fichier.com/?b0ijf36vi0wrrptp0fg2
and here is a small table of PA cities and countries. PA_town_county_table.xlsx https://1fichier.com/?0hlpk76x5jc2g0gqotea Let me know of any issue or request.
Why did the total number of vote or vote to a candidate got REDUCED at SEVERAL times during the counting evening ? what's their official explanation if they don't believe it is fraud? do they say they are human input errors?
First question: why it’s flat at republicans registered below 20%? Answer I great question, I don’t know the answer. Maybe parameters in the estimation are too scattered, but this is not a satisfactory explanation. Second question: test with Biden data or any precinct data from any other states would easily prove it one way or the other. Yes, I think so. But I don’t have access to any such data. Love to try this and other things if I have the detailed dataset.
I think Dr. Shiva s original idea of studying non party line ticket and party line ticket is an excellent idea. Add new dimension to the feature vectors. But I still think the linear slope is expected naturally. How I wish I were wrong! It would be undeniable evidence of fraud.
I think the analysis is flawed. Here is my letter to Dr. Shiva:
Hello Dr. Shiva
Always love your work for the country and with great interest I watched your presentation of data analysis on possible vote fraud in Michigan on YouTube. The data and method are great, however this time I have a different take on the conclusion of the analysis. I hope I am wrong but I don't see it.
Let r, d, and i be registered Republican, Democrat and Independent as % of all voters, respectively. i = 1-r-d. Let cr, cd and ci be % of registered Republican, Democrat and Independents who voted for Trump, respectively. (cr ~ 90%, cd ~ 3%, ci ~ 40%). And let w be % of registered Republican who voted for Trump and also choose for straight party line ticket. I have no idea what w is, maybe 70-90%? Let p and np be the % of straight party ticket (SPT) Trump votes and non-straight party (nonSPT) ticket Trump votes, of total votes, respectively. With these variables, p=rcrw; np=rcr(1-w)+dcd+(1-r-d)ci; Let delta = np - p (difference of non-SPT and SPT), then delta = (ci+cdd-cid)+(p (-ci+cr*(1-2*w)))/(cr w). This shows delta is naturally a linear function of p. Let's say cr=0.9, ci=0.4, w=0.8 (20% trump voting republican did not vote party line ticket), then the slope of delta as linear function of p is -1.3, negative as your plot of p vs. delta shows.
Actually I am very curious to know why the plot for Wayne county (Detroit) did not show this negative linear trend. That may be where the real fraud happened.
Could you please share the raw data with me? I would like to do some analysis myself to see if anything suspicious there and will share with you and public should anything worthwhile be found.
Thank you,
P.S. I am a Christian (Reformed Baptist), Conservative and Chinese American, in that order.
Hi I just checked the output csv file. It has date of birth column INTACT. I checked from top all the way to bottom. I re-created the file and re-upload it again (but it is the same content). Can you please check again https://1fichier.com/?u6w66ibi71d3hd6t60q0 do let me know if still have problem. Also do you need the other two files (5757_DataRequest 5758_...etc.) be cleaned up? I can do that if needed. Thanks.