Update data and plots.

Merge branch 'main' of https://amiok.net/gitea/W1CDN/ham-radio-licenses

# Conflicts:
#	plots/states-over-time-freey.png

README.md

@@ -8,6 +8,8 @@ This is a work in progress. It is not intended to be immediately useful for
 detailed analysis, but to act as a guide for further investigation. As we figure out
 how to slice up and caveat data, it will get more organized.
 
+There may be errors! If something looks amiss, question it!
+
 # Resources (not all integrated yet)
 - Various items in the [issue queue](https://amiok.net/gitea/W1CDN/ham-radio-licenses/issues)
 - ARRL FCC counts: https://web.archive.org/web/20150905095114/
@@ -52,3 +54,6 @@ All the data in these plots is in https://amiok.net/gitea/W1CDN/ham-radio-licens
 ![Alt text](plots/states-over-time-freey.png)
 ![Alt text](plots/cities-over-time.png)
 ![Alt text](plots/cities-over-time-freey.png)
+![Alt text](plots/actions-over-time.png)
+![Alt text](plots/actions-over-time-2-years.png)
+

bin/data-plots.R

@@ -8,6 +8,7 @@ library(plotly)
 library(htmlwidgets)
 library(lubridate)
 library(ggrepel)
+library(zoo)
 
 #### Total/State/Class ####
 # Read in total/state/class data
@@ -40,6 +41,18 @@ city_raw <- read.csv("data/us cities ham radio licenses over time.csv")
 city <- city_raw %>% mutate(Date = as.Date(Date),
                             city_label = paste0(City, ", ", State))
 
+#### License Actions ####
+ae7q_actions <- read.csv("data/ae7q-actions-scraped.csv") %>% 
+                mutate(date = as.Date(date)) %>% 
+                filter(!is.na(action)) %>% 
+                arrange(date) %>% 
+                group_by(action) %>% 
+                mutate(mean_30 = rollmean(count, k=30, fill=NA, align='right'))
+
+# Make sure all the dates are there
+#date_vec <- seq(min(ae7q_actions$date), max(ae7q_actions$date), by="days")
+#all(date_vec == unique(ae7q_actions$date))
+
 #### Plots ####
 
 ##### Total over time, y = 0 #####
@@ -283,3 +296,55 @@ ggplot(data = city,
     theme(legend.position="bottom")
 
 ggsave("plots/cities-over-time-freey.png", width = 15, height = 9)
+
+##### Actions Over Time #####
+ggplot(data = ae7q_actions,
+       aes(x = date,
+           y = count,
+           color = action)) +
+    geom_line() +
+    geom_line(data = ae7q_actions,
+              aes(x = date,
+                  y = mean_30),
+              color = "black") +
+    scale_x_date(date_breaks = "5 years",
+                 date_minor_breaks = "1 year",
+                 date_labels = "%Y") +
+    facet_wrap(~action,
+              scales = "free_y") +
+    theme_bw() +
+    labs(title = "US Amateur License Actions",
+         subtitle = "with 30-day rolling mean",
+         y = "Count",
+         x = "Date",
+         caption = "w1cdn.net; source: ae7q.com",
+         color = "Action") +
+    guides(color = "none")
+
+ggsave("plots/actions-over-time.png", width = 6, height = 4)
+
+##### Actions Over Time, last two years #####
+ggplot(data = ae7q_actions %>% filter(date >= Sys.Date() - years(2)),
+       aes(x = date,
+           y = count,
+           color = action)) +
+    geom_line() +
+    geom_line(data = ae7q_actions %>% filter(date >= Sys.Date() - years(2)),
+              aes(x = date,
+                  y = mean_30),
+              color = "black") +
+    scale_x_date(date_breaks = "6 months",
+                 date_minor_breaks = "1 months",
+                 date_labels = "%Y-%m") +
+    facet_wrap(~action,
+               scales = "free_y") +
+    theme_bw() +
+    labs(title = paste0("US Amateur License Actions since ", Sys.Date() - years(2)),
+         subtitle = "with 30-day rolling mean",
+         y = "Count",
+         x = "Date",
+         caption = "w1cdn.net; source: ae7q.com",
+         color = "Action") +
+    guides(color = "none")
+
+ggsave("plots/actions-over-time-2-years.png", width = 9, height = 6)

bin/scrape-ae7q-mass.R

@@ -0,0 +1,70 @@
+
+# Counts of license actions by date
+# Use this file to scrape a series of dates from AE7Q
+
+# Set start and end date
+date_vec <- seq(as.Date("2024-09-22"), as.Date("2024-11-26"), by="days")
+# Randomize dates we are querying
+date_vec <- sample(date_vec)
+
+dvbackup <- date_vec
+#date_vec <- date_vec[1687:7176]
+
+ae7q_list <- list()
+for(i in 1:length(date_vec)){
+    ae7q_new_url <- paste0("https://www.ae7q.com/query/list/ProcessDate.php?DATE=", date_vec[i])
+    print(ae7q_new_url)
+    
+    # Read the page
+    ae7q_new_raw <- read_html(ae7q_new_url)
+    
+    # Make sure the new license table exists first
+    if(!grepl("No license grants found issued on",  ae7q_new_raw %>% html_text())){
+        # Get tables and clean up
+        ae7q_new_tables <- ae7q_new_raw %>% 
+            html_elements(xpath = "//table") %>% 
+            html_table()
+        
+        # Find the right table by the column names
+        right_table_id <- grep(paste(c("Callsign",
+                                       "Region/ State",
+                                       "Entity Name",
+                                       "Applicant Type",
+                                       "Licensee Class",
+                                       "License Status",
+                                       "Action Type"), collapse = " "),
+                               lapply(ae7q_new_tables, function(x) paste(names(x), collapse = " ")))
+        
+        ae7q_table_new <- ae7q_new_tables[[right_table_id]]
+        
+        ae7q_list[[i]] <- ae7q_table_new %>% 
+            #mutate(across(everything(), ~na_if(., "\""))) %>% 
+            mutate(across(everything(), 
+                          ~replace(., . == "\"", NA))) %>% 
+            fill(everything()) %>% 
+            group_by(`Action Type`) %>% 
+            summarize(count = n(), .groups = "keep") %>% 
+            
+            mutate(date = date_vec[i],
+                   source = "AE7Q", source_detail = ae7q_new_url) %>% 
+            relocate(date)
+    } else {
+        ae7q_list[[i]]<- data.frame("date" = date_vec[i],
+                                    "Action Type" = NA,
+                                    "count" = NA,
+                                    "source" = "AE7Q",
+                                    "source_detail" = ae7q_new_url)
+    }
+    # Wait for random time up to 10 seconds
+    Sys.sleep(sample(1:5, 1))
+}
+
+# Combine all the data and sort by date
+a <- bind_rows(ae7q_list) %>% 
+    arrange(date) %>% 
+    filter(!is.na(date))
+
+write.csv(a, "out/ae7q-actions-scraped03.csv", row.names = F)
+
+
+

bin/scrape-license-counts.R

@@ -6,6 +6,7 @@ library(tidyr)
 # sudo crontab -e
 # 5 9 * * * su matt -c "cd /home/matt/ham-radio-licenses/; Rscript /home/matt/ham-radio-licenses/scrape-license-counts.R">/dev/null 2>&1
 
+###### ARRL ###### 
 arrl_url <- "https://www.arrl.org/fcc-license-counts"
 
 # Read the page
@@ -127,7 +128,7 @@ hamcall_table_state <- left_join(hamcall_table_state, state_codes, by = join_by(
     relocate(source_name:source_detail, .after = m)
 
 
-###### AE7Q ###### 
+###### AE7Q States ###### 
 ae7q_url <- "https://www.ae7q.com/query/stat/LicenseUSA.php"
 
 # Read the page
@@ -137,19 +138,76 @@ ae7q_raw <- read_html(ae7q_url)
 ae7q_tables <- ae7q_raw %>% 
                     html_elements(xpath = "//table") %>% 
                     html_table()
-ae7q_table_state <- ae7q_tables[[20]]
+ae7q_table_state_raw <- ae7q_tables[[20]]
 # Fix names
-names(ae7q_table_state) <- ae7q_table_state[1,]
-ae7q_table_state <- ae7q_table_state[-1,]
+names(ae7q_table_state_raw) <- ae7q_table_state_raw[1,]
+ae7q_table_state_raw <- ae7q_table_state_raw[-1,]
 
-# TODO
-# split percents out into other columns (separate_wider_delim() ?)
-# etc.
+ae7q_table_state <- ae7q_table_state_raw %>% 
+        pivot_longer(cols = -"State or Territory") %>% 
+        # remove percentages
+        mutate(value = gsub("\\s*\\([^\\)]+\\)", "", value)) %>% 
+        pivot_wider(id_cols = "State or Territory") %>% 
+        # Split states
+        separate(`State or Territory`,
+                 into = c("state_code", "state_name"),
+                 sep = " - ",
+                 fill = "right") %>% 
+        mutate(state_name = case_when(state_code == "-" ~ "Other*",
+                                      state_code == "Totals" ~ "TOTAL",
+                                      TRUE ~ state_name)) %>%
+        # Organize
+        select(c(-GeoRegion, -state_code)) %>% 
+        mutate(date = Sys.Date(),
+               ttp=NA, conditional=NA, military=NA, multiple=NA, repeater=NA,
+               gmrs=NA, source="AE7Q", source_detail=ae7q_url) %>% 
+        relocate(date, state_name, Novice, Technician, TechnicianPlus,
+                 General, Advanced, AmateurExtra, Total, ttp, conditional,
+                 Club)
 
+###### AE7Q License Actions ###### 
+ae7q_new_url <- paste0("https://www.ae7q.com/query/list/ProcessDate.php?DATE=", Sys.Date()-1)
+#ae7q_new_url <- paste0("https://www.ae7q.com/query/list/ProcessDate.php?DATE=2024-11-01")
 
+    # Read the page
+    ae7q_new_raw <- read_html(ae7q_new_url)
     
+    # Make sure the new license table exists first
+    if(!grepl("No license grants found issued on",  ae7q_new_raw %>% html_text())){
+        # Get tables and clean up
+        ae7q_new_tables <- ae7q_new_raw %>% 
+            html_elements(xpath = "//table") %>% 
+            html_table()
         
+        # Find the right table by the column names
+        right_table_id <- grep(paste(c("Callsign",
+                                       "Region/ State",
+                                       "Entity Name",
+                                       "Applicant Type",
+                                       "Licensee Class",
+                                       "License Status",
+                                       "Action Type"), collapse = " "),
+                               lapply(ae7q_new_tables, function(x) paste(names(x), collapse = " ")))
         
+        ae7q_table_new <- ae7q_new_tables[[right_table_id]]
+        
+        ae7q_sum01 <- ae7q_table_new %>% 
+            #mutate(across(everything(), ~na_if(., "\""))) %>% 
+            mutate(across(everything(), 
+                          ~replace(., . == "\"", NA))) %>% 
+            fill(everything()) %>% 
+            group_by(`Action Type`) %>% 
+            summarize(count = n(), .groups = "keep") %>% 
+            mutate(date = Sys.Date()-1,
+                   source = "AE7Q", source_detail = ae7q_new_url) %>% 
+            relocate(date)
+    } else {
+        ae7q_sum01<- data.frame("date" = Sys.Date(),
+                                    "Action Type" = NA,
+                                    "count" = NA,
+                                    "source" = "AE7Q",
+                                    "source_detail" = ae7q_new_url)
+    }
 
 
 ##### Append tables #####
@@ -167,3 +225,13 @@ write.table(hamcall_table_state, file = "out/hamcall-states-scraped.csv", sep =
             append = TRUE, quote = FALSE, 
             col.names = F, row.names = FALSE,
             na = "") 
+
+write.table(ae7q_table_state, file = "out/ae7q-states-scraped.csv", sep = ",", 
+            append = TRUE, quote = FALSE, 
+            col.names = F, row.names = FALSE,
+            na = "") 
+
+write.table(ae7q_sum01, file = "out/ae7q-actions-scraped.csv", sep = ",", 
+            append = TRUE, quote = FALSE, 
+            col.names = F, row.names = FALSE,
+            na = "")