various tweaks

README.md

@@ -40,13 +40,23 @@ Remedy: consider the conditional probabilities? But how? Or, relax assumptions u
 - [x] Download and format
 - [x] Read
 - [x] Add date of poll
-- [ ] Consider what the standards error should be 
+- [x] Consider what the standards error should be 
-- [ ] Consider how to aggregate polls?
+- [x] Consider how to aggregate polls?
   - One extreme: Just look at the most recent one
   - [x] Another extreme: Aggregate very naïvely, add up all samples together?
-- [ ] Aggregate polls?
+- [x] Aggregate polls?
-- [ ] Exclude polls older than one month?
+- [x] Exclude polls older than one month?
 - [ ] Exclude partisan polls
 - [ ] ...
 
 https://stats.stackexchange.com/questions/274211/calculating-the-probability-of-someone-winning-from-a-poll
+
+This time, the republicans win by a mile. 
+
+What is happening?
+- Trump polling ok in Minessota, 
+- No polls in Iowa yet
+- Trump polling very well in Pennsylvania, Wisconsin, Arizona, Michigan, Florida, Nevada, Georgia, North Carolina
+  - In part because of the Kennedy/West/Stein vote.
+- Biden is poling well in Colorado
+- Not that many polls yet, in general

main.go

@@ -29,10 +29,12 @@ type Poll struct {
 	SampleSize  int
 	PollResults map[string]float64
 	Date        time.Time
+	Partisan    string
 }
 
 /* Globals */
 var r = rand.New(rand.NewPCG(uint64(100), uint64(2224)))
+var dev = true
 
 /* Load data from csvs */
 func readStates() ([]State, error) {
@@ -119,8 +121,9 @@ func readStates() ([]State, error) {
 
 		poll_id := record[0]
 		state_name := record[12]
-		candidate_name := record[44]
 		end_date := record[14]
+		partisan := record[32]
+		candidate_name := record[44]
 
 		date_layout := "1/2/06"
 		parsed_date, err := time.Parse(date_layout, end_date)
@@ -150,6 +153,7 @@ func readStates() ([]State, error) {
 				SampleSize:  sample_size,
 				PollResults: make(map[string]float64),
 				Date:        parsed_date,
+				Partisan:    partisan,
 			}
 		}
 		poll.PollResults[candidate_name] = percentage
@@ -202,11 +206,12 @@ func getChanceCandidateWinsFromPoll(candidate_p float64, poll_sample_size float6
 
 /* Sample state by state */
 func sampleFromState(state State) VotesForEachParty {
-	fmt.Printf("\n\nState: %s", state.Name)
+	if dev {
-	fmt.Printf("\n\tVotes: %d", state.Votes)
+		fmt.Printf("\n\nState: %s", state.Name)
-	fmt.Printf("\n\tHistory: %s", state.PresidentialElectoralHistory)
+		fmt.Printf("\n\tVotes: %d", state.Votes)
-	// fmt.Printf("\n\tPolls: %s", state.Polls)
+		fmt.Printf("\n\tHistory: %s", state.PresidentialElectoralHistory)
-
+		// fmt.Printf("\n\tPolls: %s", state.Polls)
+	}
 	switch state.Name {
 	case "Nebraska":
 		/*
@@ -241,13 +246,14 @@ func sampleFromState(state State) VotesForEachParty {
 	default:
 		{
 			/* Just considering the base rate for the state */
-			p_republican := 0.0
+			p_baserate_republican := 0.0
 			for _, party := range state.PresidentialElectoralHistory {
 				if party == "R" {
-					p_republican++
+					p_baserate_republican++
 				}
 			}
-			p_republican = p_republican / float64(len(state.PresidentialElectoralHistory))
+			p_baserate_republican = p_baserate_republican / float64(len(state.PresidentialElectoralHistory))
+			p_republican := p_baserate_republican // if no polls
 
 			/* Considering polls */
 			var recent_polls []Poll
@@ -294,8 +300,10 @@ func sampleFromState(state State) VotesForEachParty {
 				std_poll := math.Sqrt((normalized_trump_share * normalized_biden_share) / joint_trump_biden_sample_size)
 
 				p_trump_more_votes := getProbabilityAboveX(0.5, normalized_trump_share, std_poll)
-				fmt.Printf("\n\tPoll: %+v", recent_biden_trump_poll)
+				if dev {
-				fmt.Printf("\n\t\tPoll says chance of R win: %f", p_trump_more_votes)
+					fmt.Printf("\n\tPoll: %+v", recent_biden_trump_poll)
+					fmt.Printf("\n\t\tPoll says chance of R win: %f", p_trump_more_votes)
+				}
 
 				// Update general tally
 				num_biden_votes += poll_biden_votes
@@ -308,12 +316,18 @@ func sampleFromState(state State) VotesForEachParty {
 
 				std_all_polls := math.Sqrt((aggregate_trump_share * aggregate_biden_share) / total_sample_size)
 
-				p_trump_more_votes := getProbabilityAboveX(0.5, aggregate_trump_share, std_all_polls)
+				p_republican_according_to_polls := getProbabilityAboveX(0.5, aggregate_trump_share, std_all_polls)
-				fmt.Printf("\n\tAggregating all polls naïvely says chance of R win: %f", p_trump_more_votes)
+
+				if dev {
+					fmt.Printf("\n\tAggregating all polls naïvely says chance of R win: %f", p_republican_according_to_polls)
+				}
+				p_republican = 0.75*p_republican_according_to_polls + 0.25*p_baserate_republican
 
 			}
 
-			fmt.Printf("\n\tHistorical base rate: %f", p_republican)
+			if dev {
+				fmt.Printf("\n\tHistorical base rate: %f", p_baserate_republican)
+			}
 			if r.Float64() < p_republican {
 				return VotesForEachParty{Democrats: 0, Republicans: state.Votes}
 			} else {
@@ -331,7 +345,9 @@ func simulateElection(states []State) int {
 		republican_seats += election_sample.Republicans
 	}
 
-	fmt.Printf("\n\n\n (%d) ", republican_seats)
+	if dev {
+		fmt.Printf("\n\n(%d) ", republican_seats)
+	}
 	if republican_seats >= 270 {
 		return 1
 	} else {
@@ -347,17 +363,19 @@ func main() {
 		return
 	}
 
-	n_sims := 1
+	n_sims := 10
 
 	p_republicans := 0.0
 	for i := 0; i < n_sims; i++ {
 		result := simulateElection(states)
-		fmt.Printf("Election result: %d\n", result)
+		if dev {
+			fmt.Printf("Election result: %d", result)
+		}
 		if result == 1 {
 			p_republicans++
 		}
 	}
 	p_republicans = p_republicans / float64(n_sims)
-	fmt.Printf("%% republicans: %f\n", p_republicans)
+	fmt.Printf("\n\n\n%% republicans: %f\n", p_republicans)
 
 }