Fit and interpolate data; read files

Read a file

D := READCSV("spring.csv")
D = ?
P := READPRN("spring.csv")
P = ? — As text
D := READCSV("spring.csv") D = ? P := READPRN("spring.csv") P = ?
Open this example · Download .bbm

Function	Reads
`READPRN("file")`	A text file of numbers in columns (spaces, tabs, commas or semicolons); lines that aren’t numbers, such as a header, are skipped
`READCSV("file")`	A CSV file: numbers as numbers, the rest as text
`READFILE("file", "csv")`, `READFILE("file", "delimited")`	The same, or with spaces, tabs or commas between cells
`READEXCEL("file", "Sheet1!B2:B10")`	An Excel workbook’s values (`.xlsx` or `.xls`): the first sheet’s cells, or a range

The file name is relative to the sheet’s folder, so save the sheet first.
The file is read again whenever the sheet is worked out, so the sheet follows its data.
Values come without units. Multiply a column by its unit: x := P^<0>*1 mm.

Reading files is in the desktop app; the browser can’t reach your files.

Fit a straight line

x := P^<0>*1 mm
F := P^<1>*1 N
k := slope(x, F) = ? N/mm
corr(x, F) = ? — As text
x := P^<0>*1 mm F := P^<1>*1 N k := slope(x, F) = ? N/mm corr(x, F) = ?
Open this example · Download .bbm

intercept(x, F) gives the other end of the line.

Interpolate

linterp(x, F, 5 mm) = ? N
vs := cspline(x, F)
interp(vs, x, F, 5 mm) = ? N — As text
linterp(x, F, 5 mm) = ? N vs := cspline(x, F) interp(vs, x, F, 5 mm) = ? N
Open this example · Download .bbm

linterp draws straight lines between the points.
cspline, pspline and lspline make a smooth curve through the points (with cubic, parabolic or straight ends); interp reads it at a point.
The x values must be in ascending order.

Fit a curve

regress(vx, vy, k): the least-squares polynomial of degree k, read with interp(regress(vx, vy, k), vx, vy, x).
linfit(vx, vy, F): the best coefficients for a sum of terms you choose, given as a function returning a vector: F(t) := [1; t; t^2].
For a model of your own, a solve block with minerr fits any formula.

Statistics

mean(F) = ? N
median(F) = ? N
Stdev(F) = ? N — As text
mean(F) = ? N median(F) = ? N Stdev(F) = ? N
Open this example · Download .bbm

stdev and var are for a whole population (dividing by n), Stdev and Var for a sample (by n − 1). The statistics functions include the normal distribution, histograms and random numbers.