How to decompose irregular time series?

Question

I have a time series with irregular time data from 2006 to 2020 including just the summer months (June, July, August). And I would like to get informations about trend, seasonality and mostly residuals.

• 2007-07-15 / 23.3132725761094
• 2007-07-24 / 27.5978287205735
• 2007-07-29 / 29.305232884511
• 2009-06-25 / 25.6451943453992
• 2009-06-27 / 34.6152556850167

Now,I did a complete time axis for each day over the years and merged my lst data.

library(forecast)
library(xts)

timestamps <- data$date

dataRep <- data.frame(timestamps= as.Date(c("2007-07-15","2007-07-24", "2007-07-29","2009-06-25")),
                      lst = c(23.3132725761094, 27.5978287205735, " ",25.6451943453992))

tstamp <- data.frame(x = seq(head(dataRep$timestamps,1),tail(dataRep$timestamps,1),by ="day"))
res <- merge(tstamp,dataRep,by.x="x", by.y="timestamps", all.x=TRUE)
tsData <-xts(res$lst,order.by =res$x)

Now, I wanted to apply my tsData to decompose() or slt(), but it didn't work and I got the following errors.

xts.stl <- stl(tsData,s.window="periodic")

Error: Error in na.fail.default(as.ts(x)) : missing values in object

decompose(as.ts(tsData))

Error: Error in decompose(as.ts(tsData)) : time series has no or less than 2 periods

Could someone help me please?