COVID-19 Number of Tests in Italy

Table of Contents

Menu

Introduction

This page presents some data about the number of tests and people tested for COVID-19 over time in Italy and compares them with the number of people found positive.

This page was created on <2020-08-20 Thu> and last updated on <2021-11-01 Mon>.

The source code available on the COVID-19 pages is distributed under the MIT License; the content is distributed under a Creative Commons - Attribution 4.0.

Getting data into R

We first read the data from the Civil Protection repository adding the ratio between positives and tests, computed on the same day and computed with data shifted by two days (on the assumption tests take two days to complete).

In fact data about tests is used with different semantics by different regions. Some regions reports tests with results (and the ratio new positives / tests makes sense). Other reports the number of test performed, in which case the correct ratio is between positives and tests performed some days earlier. We assume two days and report both ratios for all regions. See the following issue on GitHub for an explanation and some more details https://github.com/pcm-dpc/COVID-19/issues/577 (in Italian).

PATH="./data"
DIGITS = 4

national = read.csv(file.path(PATH, "dpc-covid19-ita-andamento-nazionale.csv"))
national$data <- as.Date(national$data)

national$nuovi_casi_testati = c(NA, diff(national$casi_testati, 1))
national$p_over_t <- round(national$nuovi_positivi / national$nuovi_casi_testati, digits = DIGITS) * 100

national$nuovi_tamponi = c(NA, diff(national$tamponi, 1))
national$p_tamponi_over_t <- round(national$nuovi_positivi / national$nuovi_tamponi, digits = DIGITS) * 100

# national$nuovi_casi_testati_2 <- c(NA, NA, head(national$nuovi_casi_testati, -2))
# national$p_over_t_2 = round(national$nuovi_positivi / national$nuovi_casi_testati_2, digits = DIGITS) * 100

# national$nuovi_tamponi_2 <- c(NA, NA, head(national$tamponi_2, -2))
# national$p_tamponi_over_t_2 = round(national$nuovi_positivi / national$nuovi_tamponi_2, digits = DIGITS) * 100

Concerning the regional level, computed columns, such as the number of people tested in a day, have to be computed after filtering, or the diif will work on values from different regions.

# evolution over time, by Region
data = read.csv(file.path(PATH, "dpc-covid19-ita-regioni.csv"))
data$data <- as.Date(data$data)

These are the columns we are interested in and their translation in English:

cols = c(
  "data",
  "nuovi_positivi",
  "nuovi_tamponi",
  "nuovi_casi_testati",
  "p_tamponi_over_t",
  "p_over_t"
)

We now define a function to ouput the last N rows of the input data frame. The real “challenge”, here, is transposing the data, to get a more natural presentation (with time progressing from left to right).

table_data <- function(df, cols, rows = 10) {
  # get the last 10 elements and the interesting columns of the dataframe
  f  <- tail(df, rows)
  rf <- f[, cols]

  # the labels in the transposed matrix are the column names of the original data.frame
  row_labels  <- colnames(rf)
  # the columns in the trasposed matrix are the dates
  col_labels  <- c("Label", format(rf$data, "%a, %b %d"))

  rft <- data.frame(row_labels, t(rf))
  colnames(rft) <- col_labels
  return(rft[-1,])
}

People Tested and Cases in Italy

Data of the last ten days

table_data(national, cols)
Label Wed, Nov 17 Thu, Nov 18 Fri, Nov 19 Sat, Nov 20 Sun, Nov 21 Mon, Nov 22 Tue, Nov 23 Wed, Nov 24 Thu, Nov 25 Fri, Nov 26
nuovi_positivi 10172 10638 10544 11555 9709 6404 10047 12448 13764 13686
nuovi_tamponi 537765 625774 538488 574812 487109 267570 689280 562505 649998 557180
nuovi_casi_testati 71521 71409 72136 75780 65752 37279 74645 78472 79663 75874
p_tamponi_over_t 1.89 1.7 1.96 2.01 1.99 2.39 1.46 2.21 2.12 2.46
p_over_t 14.22 14.9 14.62 15.25 14.77 17.18 13.46 15.86 17.28 18.04

New Cases

New cases.

## add extra space to right margin of plot within frame
par(mar=c(5, 4, 4, 6) + 0.1)

## Allow a second plot on the same graph
# par(new=TRUE)
new_cases_limits = c( min(national[national$data >= "2020-08-01", c("nuovi_positivi")]), max(national[national$data >= "2020-08-01", c("nuovi_positivi")]) )

p = plot(x = national[national$data >= "2020-08-01", c("data")], 
     y = national[national$data >= "2020-08-01", c("nuovi_positivi")], 
     type="l", lwd=6, pch=21, cex=1.5, col=c("#AA0000"),
     axes=FALSE,
     ylim=new_cases_limits,
     ylab="", xlab="")
text(x = tail(national[national$data >= "2020-08-01", c("data")], 5),
     y = tail(national[national$data >= "2020-08-01", c("nuovi_positivi")], 5),
     labels = tail(national[national$data >= "2020-08-01", c("nuovi_positivi")], 5),
     pos = 1, cex = 1, col="#AA0000")
mtext("New Cases", side=4, line=4, col="#AA0000") 
axis(4, ylim=new_cases_limits, las=1)

grid(p, col = "black", lty = "dotted")

# x-axis
dates = national[national$data >= "2020-08-01", c("data")]
axis.Date(1, at=seq(min(dates), max(dates), by="week"), format="%b %d", las=2)
mtext("Day", side=1, line=2.5)

## Add Legend
legend("topleft", legend = c("Tests", "New Cases"),
       text.col = c("#3B3176", "#AA0000"), pch= c(15, 17), col=c("#3B3176", "#AA0000"))

new_cases_italia.png

New Cases Tested

plot(x = national[national$data >= "2020-08-01", c("data")], 
     y = national[national$data >= "2020-08-01", c("nuovi_casi_testati")], 
     type="l", lwd=6, pch=16, cex=2.5, col=c("#3B3176"))
text(x = tail(national[national$data >= "2020-08-01", c("data")], 1),
     y = tail(national[national$data >= "2020-08-01", c("nuovi_casi_testati")], 1),
     labels = tail(national[national$data >= "2020-08-01", c("nuovi_casi_testati")], 1),
     pos = 4, cex = 1.2, col=c("#3B3176"))
 grid(col="black")

tests_italia.png

Number of Tests and New Cases Tested

Plot new cases and tests together. (Solution taken from How can I plot with 2 different y-axes? on Stack Overflow.)

## add extra space to right margin of plot within frame
par(mar=c(5, 4, 4, 6) + 0.1)

## Plot first set of data and draw its axis
tests_limits = c( min(national[national$data >= "2020-08-01", c("nuovi_casi_testati")]), max(national[national$data >= "2020-08-01", c("nuovi_casi_testati")]) )
plot(x = national[national$data >= "2020-08-01", c("data")], 
     y = national[national$data >= "2020-08-01", c("nuovi_casi_testati")], 
     type="l", lwd=6, pch=11, cex=1.5, col=c("#3B3176"),
     axes=FALSE,
     ylim=tests_limits,
     ylab="", xlab="")
text(x = tail(national[national$data >= "2020-08-01", c("data")], 1),
     y = tail(national[national$data >= "2020-08-01", c("nuovi_casi_testati")], 1),
     labels = tail(national[national$data >= "2020-08-01", c("nuovi_casi_testati")], 1),
     pos = 4, cex = 1, col=c("#3B3176"))
mtext("Number of Tests", side=2, col="#3B3176", line=4) 
axis(2, ylim=tests_limits, col="black", las=1)  
box()

## Allow a second plot on the same graph
par(new=TRUE)
new_cases_limits = c( min(national[national$data >= "2020-08-01", c("nuovi_positivi")]), max(national[national$data >= "2020-08-01", c("nuovi_positivi")]) )

p = plot(x = national[national$data >= "2020-08-01", c("data")], 
     y = national[national$data >= "2020-08-01", c("nuovi_positivi")], 
     type="l", lwd=6, pch=21, cex=1.5, col=c("#AA0000"),
     axes=FALSE,
     ylim=new_cases_limits,
     ylab="", xlab="")
text(x = tail(national[national$data >= "2020-08-01", c("data")], 1),
     y = tail(national[national$data >= "2020-08-01", c("nuovi_positivi")], 1),
     labels = tail(national[national$data >= "2020-08-01", c("nuovi_positivi")], 1),
     pos = 4, cex = 1, col="#AA0000")
mtext("New Cases", side=4, line=4, col="#AA0000") 
axis(4, ylim=new_cases_limits, las=1)

grid(p, col = "black", lty = "dotted")

# x-axis
dates = national[national$data >= "2020-08-01", c("data")]
axis.Date(1, at=seq(min(dates), max(dates), by="week"), format="%b %d", las=2)
mtext("Day", side=1, line=2.5)

## Add Legend
legend("topleft", legend = c("Tests", "New Cases"),
       text.col = c("#3B3176", "#AA0000"), pch= c(15, 17), col=c("#3B3176", "#AA0000"))

tests_and_new_cases_italia.png

Positive/Number of Tests

Here we plot the number of positive people over tests performed. The standard measurement is the ratio between positive and tests performed (shown in blue). The way I understand it is that this number also includes tests performed on people already diagnosed and recovered.

The second graph, in red, shows the ration of positive over new people tested, that is, of all the people not yet diagnosed, how many resulted positive?

plot(national$p_over_t ~ national$data, type="o", lwd=3, pch=21, col="#ff0000", main="Positive over Tests", xlab="Date", ylab="Percentage")
text(y = tail(national, 1)$p_over_t, x = tail(national, 1)$data, lab = paste(tail(national, 1)$p_over_t, "%", sep=""), pos=4, col="#ff0000", cex=1.3)

# Second plot with Positive over tests
p = lines(national$p_tamponi_over_t ~ national$data, type="o", lwd=3, pch=21, col="#000088", xlab="Date", ylab="Percentage")
text(y = tail(national, 1)$p_tamponi_over_t, x = tail(national, 1)$data, lab = paste(tail(national, 1)$p_tamponi_over_t, "%", sep=""), pos=4, col="#000088", cex=1.3)

## Add Legend
grid(col="black")
legend("bottomleft", legend = c("Positive over new People Tested", "Positive over Tests Performed"),
       text.col = c("#ff0000", "#000088"), pch= c(15, 17), col=c("#AA0000", "#000088"))

positive_over_tests_italia.png

People Tested and Cases in Trentino

region <- subset(data, denominazione_regione == "P.A. Trento")

region$nuovi_casi_testati = c(NA, diff(region$casi_testati, 1))

region$p_over_t <- round(region$nuovi_positivi / region$nuovi_casi_testati, digits = DIGITS) * 100
region$nuovi_casi_testati_2 = c(NA, NA, diff(region$casi_testati, 2))
region$p_over_t_2 = round(region$nuovi_positivi / region$nuovi_casi_testati_2, digits = DIGITS) * 100
region$nuovi_casi_testati_2 <- c(NA, NA, head(region$nuovi_casi_testati, -2))
region$p_over_t_2 = round(region$nuovi_positivi / region$nuovi_casi_testati_2, digits = DIGITS) * 100

region$nuovi_tamponi = c(NA, diff(region$tamponi, 1))
region$p_tamponi_over_t <- round(region$nuovi_positivi / region$nuovi_tamponi, digits = DIGITS) * 100
region$nuovi_tamponi_2 <- c(NA, NA, head(region$tamponi_2, -2))
region$p_tamponi_over_t_2 = round(region$nuovi_positivi / region$nuovi_tamponi_2, digits = DIGITS) * 100

table_data(region, cols)
Label Wed, Nov 17 Thu, Nov 18 Fri, Nov 19 Sat, Nov 20 Sun, Nov 21 Mon, Nov 22 Tue, Nov 23 Wed, Nov 24 Thu, Nov 25 Fri, Nov 26
nuovi_positivi 112 95 102 164 96 71 101 151 177 138
nuovi_tamponi 7642 10905 8569 9107 7175 4830 12236 7986 11176 8239
nuovi_casi_testati 504 507 515 600 510 506 520 475 514 482
p_tamponi_over_t 1.47 0.87 1.19 1.8 1.34 1.47 0.83 1.89 1.58 1.67
p_over_t 22.22 18.74 19.81 27.33 18.82 14.03 19.42 31.79 34.44 28.63

People Tested and Cases in Liguria

region <- subset(data, denominazione_regione == "Liguria")

region$nuovi_casi_testati = c(NA, diff(region$casi_testati, 1))

region$p_over_t <- round(region$nuovi_positivi / region$nuovi_casi_testati, digits = DIGITS) * 100
region$nuovi_casi_testati_2 = c(NA, NA, diff(region$casi_testati, 2))

region$nuovi_tamponi = c(NA, diff(region$tamponi, 1))
region$p_tamponi_over_t <- round(region$nuovi_positivi / region$nuovi_tamponi, digits = DIGITS) * 100

table_data(region, cols)
Label Wed, Nov 17 Thu, Nov 18 Fri, Nov 19 Sat, Nov 20 Sun, Nov 21 Mon, Nov 22 Tue, Nov 23 Wed, Nov 24 Thu, Nov 25 Fri, Nov 26
nuovi_positivi 271 315 351 337 313 137 350 412 460 460
nuovi_tamponi 13910 16405 13869 16309 11723 5184 19789 15198 16380 14326
nuovi_casi_testati 1926 1964 1526 2384 1833 585 2240 2022 1879 1818
p_tamponi_over_t 1.95 1.92 2.53 2.07 2.67 2.64 1.77 2.71 2.81 3.21
p_over_t 14.07 16.04 23.0 14.14 17.08 23.42 15.62 20.38 24.48 25.3

People Tested and Cases in Veneto

region <- subset(data, denominazione_regione == "Veneto")

region$nuovi_casi_testati = c(NA, diff(region$casi_testati, 1))
region$p_over_t <- round(region$nuovi_positivi / region$nuovi_casi_testati, digits = DIGITS) * 100

region$nuovi_tamponi = c(NA, diff(region$tamponi, 1))
region$p_tamponi_over_t <- round(region$nuovi_positivi / region$nuovi_tamponi, digits = DIGITS) * 100

table_data(region, cols)
Label Wed, Nov 17 Thu, Nov 18 Fri, Nov 19 Sat, Nov 20 Sun, Nov 21 Mon, Nov 22 Tue, Nov 23 Wed, Nov 24 Thu, Nov 25 Fri, Nov 26
nuovi_positivi 1435 1603 1283 1928 1261 870 1632 1931 2066 2036
nuovi_tamponi 85207 117123 95657 93439 83981 33881 123768 90592 122438 91448
nuovi_casi_testati 4431 4762 5039 3993 3612 2819 4892 4770 5566 4729
p_tamponi_over_t 1.68 1.37 1.34 2.06 1.5 2.57 1.32 2.13 1.69 2.23
p_over_t 32.39 33.66 25.46 48.28 34.91 30.86 33.36 40.48 37.12 43.05

People Tested and Cases in Lombardia

region <- subset(data, denominazione_regione == "Lombardia")

region$nuovi_casi_testati = c(NA, diff(region$casi_testati, 1))
region$p_over_t <- round(region$nuovi_positivi / region$nuovi_casi_testati, digits = DIGITS) * 100

region$nuovi_tamponi = c(NA, diff(region$tamponi, 1))
region$p_tamponi_over_t <- round(region$nuovi_positivi / region$nuovi_tamponi, digits = DIGITS) * 100

table_data(region, cols)
Label Wed, Nov 17 Thu, Nov 18 Fri, Nov 19 Sat, Nov 20 Sun, Nov 21 Mon, Nov 22 Tue, Nov 23 Wed, Nov 24 Thu, Nov 25 Fri, Nov 26
nuovi_positivi 1858 1705 1735 1930 1431 662 1668 2207 2302 2209
nuovi_tamponi 106695 135080 105057 118279 106555 41291 153509 113920 143541 115036
nuovi_casi_testati 13664 12972 12920 14603 12958 5392 14767 14859 14208 14129
p_tamponi_over_t 1.74 1.26 1.65 1.63 1.34 1.6 1.09 1.94 1.6 1.92
p_over_t 13.6 13.14 13.43 13.22 11.04 12.28 11.3 14.85 16.2 15.63