Commit: ead5c3b4e42a711babe2a69ef5f59a641abf6e4a - myprojects-hg-reborn (hg)

Commit MetaInfo

• Revision: ead5c3b4e42a711babe2a69ef5f59a641abf6e4a (tree)
• Time: 2021-01-17 07:07:42
• Author: Lorenzo Isella <lorenzo.isella@gmai...>
• Commiter: Lorenzo Isella

Log Message

I completed the code with all the ranking stuff.

Change Summary

• modified: R-codes/process_baci.R

Incremental Difference

diff -r 0ecd8ac171bf -r ead5c3b4e42a R-codes/process_baci.R
--- a/R-codes/process_baci.R	Fri Jan 15 16:29:29 2021 +0100
+++ b/R-codes/process_baci.R	Sat Jan 16 23:07:42 2021 +0100
@@ -1,7 +1,7 @@
 rm(list=ls())
 
 
-## last saved on  Time-stamp: "2021-01-15 16:22:40 lorenzo"
+## last saved on  Time-stamp: "2021-01-16 23:06:49 lorenzo"
 
   
 
@@ -15,6 +15,9 @@
 
 
 
+
+n_rank <- 20 ### number of values for the discretized ranking
+
 eu_list <- c("BE", "BG","CZ", "DK", "DE", "EE", "IE", "GR",
                    "ES", "FR", "IT", "CY", "LV", "LT", "LU", "HU", "MT",
              "NL", "AT","PL", "PT", "RO", "SI", "SK", "FI","SE", "HR" ) %>%
@@ -339,12 +342,42 @@
     full_join(y=a6) %>%
     full_join(y=a7) %>%
     full_join(y=a8) %>%
-    round_all(3) %>%
+    ## round_all(3) %>%
     left_join(y=prod_codes, by=c("k"="code")) 
 
 
 
 
+## test_a2 <- a2 %>%
+##     mutate(N_rank=dense_rank(N_eu_imports)) %>%
+##     arrange(N_rank)
+
+
+## test_a2_round <- a2 %>%
+##     round_all(3) %>% 
+##     mutate(N_rank=dense_rank(N_eu_imports)) %>%
+##     arrange(N_rank)
+
+
+
+
+## test_a4 <- a4 %>%
+##     mutate(HHI_rank=dense_rank(HHI_EU_imports)) %>%
+##     arrange(HHI_rank)
+
+
+## test_a4_round <- a4 %>%
+##     round_all(3) %>% 
+##     mutate(HHI_rank=dense_rank(HHI_EU_imports)) %>%
+##     arrange(HHI_rank)
+
+
+
+
+
+
+
+
 ## df_tot_round <- a1 %>%
 ##     full_join(y=a2) %>%
 ##     full_join(y=a3) %>%
@@ -378,7 +411,8 @@
 
 extra_eu_imports <- df_imports_raw %>%
     select(k, i_value) %>%
-    group_by(k) %>%
+    group_by(k) %>%  ## I just consider the value of the imports from
+                     ### extra EU for every product line k  
     summarise(extra_EU_import_value=sum(i_value, na.rm=T))
 
 
@@ -394,7 +428,7 @@
     mutate(rank_N_eu=dense_rank(N_eu_imports),
            rank_HHI_EU=dense_rank(desc(HHI_EU_imports)),
            rank_balance=dense_rank(balance_over_imports),
-           rank_price_diff=dense_rank(desc(import_price_difference)),
+           rank_price_diff=dense_rank(desc(import_price_difference))
            ) %>%
     rowwise %>% 
     mutate(mean_ranking=mean(c_across(rank_N_eu:rank_price_diff),na.rm=T)) %>%
@@ -440,11 +474,15 @@
     select(-q) %>% 
     filter(i %in% codes_eu$country_code, j %in% codes_eu$country_code,
            t==max(t))  %>%
-    filter(k %in% unique(df_rank$k)) %>%
+    filter(k %in% df_rank$k) %>%
     group_by(k) %>%
     summarise(number_EU_providers=length(unique(i)),
               value_intra_EU_exports=sum(v, na.rm=T)) %>%
-    ungroup
+    ungroup  ### it is very important that I use the unique number of EU
+### exporters (unique(i)): in general they are repeated because, for a given
+                                        # product k, one exporter can export to different EU countries.
+###Instead, I can just sum the value of the exports because I do not need
+### to consider whether the exporter exports to one or more EU countrues
 
 
 
@@ -667,6 +705,21 @@
 save_excel(df_hs2, "share_hs2.xlsx")
 
 
+
+df_rank_final <- df_rank %>%
+    mutate(rank_N_eu_discrete=ntile(N_eu_imports, n_rank),
+           rank_HHI_EU_discrete=ntile(desc(HHI_EU_imports), n_rank),
+           rank_balance_discrete=ntile(balance_over_imports,n_rank),
+           rank_price_diff_discrete=ntile(desc(import_price_difference),n_rank)
+           ) %>%
+        rowwise %>% 
+    mutate(mean_ranking_discrete=mean(c_across(rank_N_eu_discrete:rank_price_diff_discrete),na.rm=T)) %>%
+    ungroup 
+
+    
+save_excel(df_rank_final, "final_ranking_complete.xlsx")
+
+
 print("So far so good")