STALE_VALUES_ROUND

This function detects “stale” or repeating sensor values that occur when an instrument becomes stuck.

A point is marked as stale if it belongs to a contiguous sequence (of length at least window) of readings that are identical when rounded to the specified number of decimal places.

Excel Usage

=STALE_VALUES_ROUND(x, window, decimals, mark)

x (list[list], required): Numeric data values.
window (int, optional, default: 6): Number of consecutive identical elements to be considered stale.
decimals (int, optional, default: 3): Number of decimal places to round to before comparing.
mark (str, optional, default: “tail”): How much of the sequence to mark True. One of ‘tail’, ‘end’, or ‘all’.

Returns (list[list]): 2D list of booleans (True if stale), or an error string.

Example 1: Detect repeating values over a window

Inputs:

x	window	decimals	mark
1.001	3	2	all
1.001
1.001
1.001
2

Excel formula:

=STALE_VALUES_ROUND({1.001;1.001;1.001;1.001;2}, 3, 2, "all")

Expected output:

Result
true
true
true
true
false

Example 2: Detect stale values with tail marking mode

Inputs:

x	window	decimals	mark
5	3	0	tail
5
5
6

Excel formula:

=STALE_VALUES_ROUND({5;5;5;6}, 3, 0, "tail")

Expected output:

Result
false
true
true
false

Example 3: Detect stale values with end marking mode

Inputs:

x	window	decimals	mark
2.2	3	1	end
2.2
2.2
2.2
2.2

Excel formula:

=STALE_VALUES_ROUND({2.2;2.2;2.2;2.2;2.2}, 3, 1, "end")

Expected output:

Result
false
false
true
true
true

Example 4: Handle scalar stale-value input

Inputs:

x	window	decimals	mark
7.5	3	2	all

Excel formula:

=STALE_VALUES_ROUND(7.5, 3, 2, "all")

Expected output:

false

Python Code

Show Code

import pandas as pd
from pvanalytics.quality.gaps import stale_values_round as result_func

def stale_values_round(x, window=6, decimals=3, mark='tail'):
    """
    Identify stale or stuck-sensor values by rounding and checking for repeating sequences.

    See: https://pvanalytics.readthedocs.io/en/stable/generated/pvanalytics.quality.gaps.stale_values_round.html

    This example function is provided as-is without any representation of accuracy.

    Args:
        x (list[list]): Numeric data values.
        window (int, optional): Number of consecutive identical elements to be considered stale. Default is 6.
        decimals (int, optional): Number of decimal places to round to before comparing. Default is 3.
        mark (str, optional): How much of the sequence to mark True. One of 'tail', 'end', or 'all'. Valid options: Tail, End, All. Default is 'tail'.

    Returns:
        list[list]: 2D list of booleans (True if stale), or an error string.
    """
    try:
        def flatten_num(data):
            if not isinstance(data, list): return [float(data)]
            flat = []
            for row in data:
                row = row if isinstance(row, list) else [row]
                for val in row:
                    if val == "": flat.append(float('nan'))
                    else: flat.append(float(val))
            return flat

        val_list = flatten_num(x)

        if len(val_list) == 0:
            return "Error: input array cannot be empty"

        series = pd.Series(val_list)

        w = int(window) if window is not None else 6
        dec = int(decimals) if decimals is not None else 3
        m = str(mark) if mark is not None else "tail"

        if m not in ["tail", "end", "all"]:
            return "Error: mark must be 'tail', 'end', or 'all'"

        res = result_func(series, window=w, decimals=dec, mark=m)

        return [[bool(v)] for v in res]
    except Exception as e:
        return f"Error: {str(e)}"

Online Calculator

x *

Numeric data values.

window

Number of consecutive identical elements to be considered stale.

decimals

Number of decimal places to round to before comparing.

mark

How much of the sequence to mark True. One of 'tail', 'end', or 'all'.