BAD_Analytics/R/bad_proj_analytics.Rmd

---
title: "BAD_Proj_Analytics"
author: "Marcin Kostrzewski"
date: "13/05/2019"
output: pdf_document
---

```{r, echo=FALSE}
con <- DBI::dbConnect(odbc::odbc(),
                      driver = "SQL Server",
                      database = "dbad_flights",
                      UID    = "dbad_s444409",
                      PWD    = "qAOKvlYA6e",
                      server = "mssql-2016.labs.wmi.amu.edu.pl",
                      port = 5432)
```

## Jakie było średnie opóźnienie przylotu?

```{sql, connection=con, tab.cap = NA}
SELECT CAST(AVG(arr_delay) AS NUMERIC(30,3)) AS 'Average delay (minutes)'
FROM Flight_delays
WHERE arr_delay IS NOT NULL;
```

## Jakie było maksymalne opóźnienie przylotu?

```{sql, connection=con, tab.cap = NA}
SELECT CAST(MAX(arr_delay)/60 AS NUMERIC(30,3)) AS 'Max delay (hours)'
FROM Flight_delays
WHERE arr_delay IS NOT NULL;
```

## Który lot miał największe opóźnienie przylotu?

```{sql, connection=con, tab.cap = NA}
SELECT carrier AS 'Carrier',
       origin_city_name AS 'Origin',
       dest_city_name AS 'Destination',
       fl_date AS 'Date',
       arr_delay AS 'Delay (minutes)'
FROM   Flight_delays
WHERE  arr_delay = (SELECT MAX(arr_delay)
                    FROM   Flight_delays
                    WHERE  arr_delay IS NOT NULL);
```

## Które dni tygodnia są najgorsze do podróżowania?

```{sql, connection=con, tab.cap = NA}
SELECT   CASE WHEN day_of_week = 1 THEN 'Monday'
              WHEN day_of_week = 2 THEN 'Tuesday'
              WHEN day_of_week = 3 THEN 'Wednesday'
              WHEN day_of_week = 4 THEN 'Thursday'
              WHEN day_of_week = 5 THEN 'Friday'
              WHEN day_of_week = 6 THEN 'Saturday'
              WHEN day_of_week = 7 THEN 'Sunday'
         END AS 'Day',
         AVG(arr_delay) AS 'Average Delay (minutes)'
FROM     Flight_delays
GROUP BY day_of_week
ORDER BY AVG(arr_delay) DESC;
```

## Które linie lotnicze latające z San Francisco (SFO) mają najmniejsze opóźnienia przylotu?

```{sql, connection=con, tab.cap = NA}
SELECT   F1.carrier AS 'Carrier',
         (SELECT   AVG(F2.arr_delay) AS 'avg_delay'
          FROM     Flight_delays F2
          WHERE    F1.carrier = F2.carrier
          GROUP BY F2.carrier) AS 'Delay (minutes)'
FROM     Flight_delays F1
WHERE    F1.origin_city_name LIKE 'San Francisco%'
GROUP BY F1.carrier
ORDER BY "Delay (minutes)" ASC;
```

*Pojawiające się w tabeli wartości ujemne oznaczają, że średnio samoloty lądowały wcześniej, niż przewidziano, czyli były przyśpieszone.*

## Jaka część linii lotniczych ma regularne opóźnienia, tj. jej lot ma średnio co najmniej 10 min. opóźnienia?

```{sql, connection=con, tab.cap = NA, echo=FALSE}
SELECT (SELECT COUNT(*)
        FROM (SELECT   F1.carrier,
                       (SELECT AVG(arr_delay)
                        FROM   Flight_delays F2
                        WHERE  F1.carrier=F2.carrier
                        GROUP BY F2.carrier
                        HAVING AVG(F1.arr_delay)>10) AS 'avg_delay'
              FROM     Flight_delays F1
              GROUP BY F1.carrier) AS t
              WHERE    t.avg_delay IS NOT NULL) / CAST((SELECT COUNT(*)
                                                        FROM   (SELECT   COUNT(*) AS 'count'
                                                                FROM     Flight_delays
                                                                GROUP BY carrier) AS T) AS FLOAT) AS 'Part of continuous delays';
```

## Jak opóźnienia wylotów wpływają na opóźnienia przylotów?
```{r, con, echo=FALSE}
data <- DBI::dbGetQuery(con, "SELECT dep_delay,
       arr_delay
FROM   Flight_delays
WHERE  dep_delay IS NOT NULL AND arr_delay IS NOT NULL;")

library(knitr)
res <- cor(data, use = "all", method = "pearson")
kable(res[2:2])
```