trend narrative fixed for time periods and shorter time series

R/helpers.R

@@ -264,6 +264,14 @@ pytd_volume <- function(
     previous_year <- lubridate::year(py_date)
   }
 
+  df <- df %>%
+    dplyr::mutate(year = lubridate::year(base::get(date))) %>%
+    dplyr::filter(year == previous_year,
+                  base::get(date) <= py_date)
+
+  # return NA rather than 0 if there are no records for previous year
+  if (nrow(df) == 0) return(NA)
+
   py_volume <- df %>%
     dplyr::mutate(year = lubridate::year(base::get(date))) %>%
     dplyr::filter(year == previous_year,
@@ -413,7 +421,10 @@ get_trend_outliers <- function(
     ) %>%
     dplyr::ungroup() %>%
     dplyr::select(-data) %>%
-    dplyr::arrange(-abs_change)
+    dplyr::arrange(-abs_change) %>%
+    dplyr::filter(!is.na(change))
+
+  if (nrow(table) == 0) return(NULL)
 
   if (summarization %in% c("sum", "count")) {
 

R/narrate_trend.R

@@ -241,76 +241,83 @@ narrate_trend <- function(
   narrative_name <- glue::glue("{timeframe_curr} vs {timeframe_prev}")
 
   change <- round(total_curr_raw - total_prev_raw, 1)
-  change_p <- paste(round((total_curr_raw/total_prev_raw - 1)*100, 1), "%")
-  trend <- ifelse(change > 0, "increase", "decrease")
 
-  if (format_numbers == TRUE) {
-    total_curr <- format_num(total_curr_raw, decimals = 2)
-    total_prev <- format_num(total_prev_raw, decimals = 2)
-    change <- format_num(change, decimals = 2)
-  } else {
-    total_curr <- total_curr_raw
-    total_prev <- total_prev_raw
-    change <- format_num(change, decimals = 2)
-  }
+  if (!is.na(change)) {
 
-  # Sum/Count ---------------------------------------------------------------
-  if (summarization %in% c("sum", "count")) {
+    change_p <- paste(round((total_curr_raw/total_prev_raw - 1)*100, 1), "%")
+    trend <- ifelse(change > 0, "increase", "decrease")
 
-    narrative_total <- glue::glue(
-      template_total,
-      .transformer = collapse_transformer(sep = collapse_sep, last = collapse_last)
-    )
+    if (format_numbers == TRUE) {
+      total_curr <- format_num(total_curr_raw, decimals = 2)
+      total_prev <- format_num(total_prev_raw, decimals = 2)
+      change <- format_num(change, decimals = 2)
+    } else {
+      total_curr <- total_curr_raw
+      total_prev <- total_prev_raw
+      change <- format_num(change, decimals = 2)
+    }
+
+    # Sum/Count ---------------------------------------------------------------
+    if (summarization %in% c("sum", "count")) {
 
-    narrative <- list(narrative_total) %>%
-      rlang::set_names(narrative_name)
-
-    variables <- list(
-      list(
-        narrative_total = narrative_total,
-        template_total = template_total,
-        measure = measure,
-        total_curr = total_curr,
-        total_curr_raw = total_curr_raw,
-        total_prev = total_prev,
-        total_prev_raw = total_prev_raw,
-        timeframe_curr = timeframe_curr,
-        timeframe_prev = timeframe_prev,
-        change = change,
-        change_p = change_p,
-        trend = trend
+      narrative_total <- glue::glue(
+        template_total,
+        .transformer = collapse_transformer(sep = collapse_sep, last = collapse_last)
       )
-    ) %>%
-      rlang::set_names(narrative_name)
 
-    # Average ------------------------------------------------------------
-  } else if (summarization == "average") {
+      narrative <- list(narrative_total) %>%
+        rlang::set_names(narrative_name)
 
-    narrative_average <- glue::glue(
-      template_average,
-      .transformer = collapse_transformer(sep = collapse_sep, last = collapse_last)
-    )
+      variables <- list(
+        list(
+          narrative_total = narrative_total,
+          template_total = template_total,
+          measure = measure,
+          total_curr = total_curr,
+          total_curr_raw = total_curr_raw,
+          total_prev = total_prev,
+          total_prev_raw = total_prev_raw,
+          timeframe_curr = timeframe_curr,
+          timeframe_prev = timeframe_prev,
+          change = change,
+          change_p = change_p,
+          trend = trend
+        )
+      ) %>%
+        rlang::set_names(narrative_name)
 
-    narrative <- list(narrative_average) %>%
-      rlang::set_names(narrative_name)
-
-    variables <- list(
-      list(
-        narrative_average = narrative_average,
-        template_average = template_average,
-        measure = measure,
-        total_curr = total_curr,
-        total_curr_raw = total_curr_raw,
-        total_prev = total_prev,
-        total_prev_raw = total_prev_raw,
-        timeframe_curr = timeframe_curr,
-        timeframe_prev = timeframe_prev,
-        change = change,
-        change_p = change_p,
-        trend = trend
+      # Average ------------------------------------------------------------
+    } else if (summarization == "average") {
+
+      narrative_average <- glue::glue(
+        template_average,
+        .transformer = collapse_transformer(sep = collapse_sep, last = collapse_last)
       )
-    ) %>%
-      rlang::set_names(narrative_name)
+
+      narrative <- list(narrative_average) %>%
+        rlang::set_names(narrative_name)
+
+      variables <- list(
+        list(
+          narrative_average = narrative_average,
+          template_average = template_average,
+          measure = measure,
+          total_curr = total_curr,
+          total_curr_raw = total_curr_raw,
+          total_prev = total_prev,
+          total_prev_raw = total_prev_raw,
+          timeframe_curr = timeframe_curr,
+          timeframe_prev = timeframe_prev,
+          change = change,
+          change_p = change_p,
+          trend = trend
+        )
+      ) %>%
+        rlang::set_names(narrative_name)
+    }
+  } else {
+    narrative <- list()
+    variables <- list()
   }
 
   # Adding Date as dimension and leaving last two years only
@@ -558,6 +565,8 @@ narrate_trend <- function(
   }
 
   # Output ------------------------------------------------------------------
+  if (length(narrative) == 0) return(NULL)
+
   if (return_data == TRUE) {
     return(variables)
   }

README.md

@@ -135,26 +135,6 @@ narrative_enhanced <- enhance_narrative(narrative_one)
 cat(narrative_enhanced)
 ```
 
-The Total Sales generated across all Regions amounted to a staggering
-38790478.4. Upon closer examination, it was found that the Outlying
-Regions, namely NA and EMEA, contributed significantly to this
-impressive figure. NA accounted for 18079736.4, which translates to
-46.6%, while EMEA contributed 13555412.7, which is 34.9% of the Total
-Sales.
-
-Further analysis of the sales data revealed that within NA, the
-significant Products that contributed to these sales figures were Food &
-Beverage, generating 7392821 or 40.9%, followed by Electronics, which
-generated 3789132.7 or 21%. Similarly, within EMEA, the significant
-Products were Food & Beverage, generating 5265113.2 or 38.8%, followed
-by Electronics, which generated 3182803.4 or 23.5%.
-
-It is noteworthy to mention that the Outlying Products that contributed
-to the Total Sales were primarily Food & Beverage, generating 15543469.7
-or 40.1%, followed by Electronics, which generated 8608962.8 or 22.2%.
-These sales figures are a testament to the strong performance of the
-Food & Beverage category across all Regions.
-
 ### Translation
 
 Translate you text using `translate_narrative()` function, specify
@@ -165,25 +145,6 @@ translation <- translate_narrative(narrative_enhanced, language = "Czech")
 cat(translation)
 ```
 
-Celkové tržby vygenerované ve všech oblastech dosáhly ohromujícího čísla
-38790478.4. Po bližším prozkoumání bylo zjištěno, že odlehlé oblasti,
-konkrétně NA a EMEA, významně přispěly k tomuto úžasnému číslu. NA
-představuje 18079736.4, což představuje 46,6%, zatímco EMEA přispěla
-13555412.7, což je 34,9% celkových tržeb.
-
-Další analýza prodejních dat odhalila, že v rámci NA jsou významné
-produkty, které přispěly k těmto prodejním číslům, potraviny a nápoje,
-které vygenerovaly 7392821 nebo 40,9%, následované elektronikou, která
-vygenerovala 3789132.7 nebo 21%. Podobně v rámci EMEA jsou významné
-produkty potraviny a nápoje, které vygenerovaly 5265113.2 nebo 38,8%,
-následované elektronikou, která vygenerovala 3182803.4 nebo 23,5%.
-
-Je důležité zmínit, že odlehlé produkty, které přispěly k celkovým
-tržbám, byly především potraviny a nápoje, které vygenerovaly 15543469.7
-nebo 40,1%, následované elektronikou, která vygenerovala 8608962.8 nebo
-22,2%. Tyto prodejní čísla jsou důkazem silného výkonu kategorie
-potravin a nápojů ve všech oblastech.
-
 ### Summarization
 
 If your output is too verbose you can summarize it with
@@ -194,13 +155,6 @@ summarization <- summarize_narrative(narrative_enhanced)
 cat(summarization)
 ```
 
-Total sales were 38790478.4, with significant contributions from North
-America (46.6%) and EMEA (34.9%). Food & Beverage was the top-selling
-product in both regions, generating 40.9% and 38.8% of sales,
-respectively. Electronics followed with 21% and 23.5% of sales. Food &
-Beverage and Electronics were also the top-selling products overall,
-contributing 40.1% and 22.2% to total sales, respectively.
-
 # Python
 
 Here are some basic Python examples, for more details visit [pynarrator

docs/pkgdown.yml

@@ -8,7 +8,7 @@ articles:
   formatting: formatting.html
   templates: templates.html
   trend: trend.html
-last_built: 2023-06-25T17:52Z
+last_built: 2023-07-03T14:29Z
 urls:
   reference: https://denisabd.github.io/narrator/reference
   article: https://denisabd.github.io/narrator/articles

package_prep.R

@@ -28,7 +28,6 @@ pkgdown::build_site(lazy = FALSE, new_process = TRUE)
 
 # devtools checks and test ---------------------------------------------------------
 devtools::test()
-
 devtools::check()
 
 # code coverage