(from version 3.6.5) To correct a chronoamperogram that is s step
function distorted by film loss, according to the method in ref
 load it and run the
film_decay command. The
INFO: currently 1 partitions Partition data (f): fixed size division (j): jump-based estimation (s): save partition (l): load previously saved partition (q): cancel (y,o,mid): accept... as SOAS is trying to identify the different intervals.
f if the data whose film decay you wish to correct
consists of intervals that have the same size.
j, for the typical case where steps are of random
size, but well separated by a ``jump'' in current. After you type
j, the terminal says:
INFO: Starting auto partitioning INFO: currently 4 partitions Auto-partition data (q,o,y, mid) to accept Jump partition parameters (i) minimum interval size: 50 (a) averaging over 10 elements (j) jump factor: 2.000 (m) minimum delta: 0.2300E-07
Soas uses a rather crude algorithm to detect those jumps: it will
detect a new interval only
i points (default is
after the previous one, if the value of the difference between two
consecutive points is larger than the minimum delta (parameter
m) and at least
j times (default
j=2) larger than
the average difference over the last
a elements (default
a=10). Of course the four parameters can be set to different
This screen capture illustrates the result:
It is strongly advised to remove spikes on the data before using the ``jump-based partitioning'' scheme.
Accept the partition by typing
by clicking the middle button of the mouse; you are back to the
After the particion is accepted, the terminal says:
INFO: currently 4 partitions Partition data (f): fixed size division (j): jump-based estimation (s): save partition (l): load previously saved partition (q): cancel (y,o,mid): acceptIt is possible to save the current partition and reload it from a file using
l. The partition is saved as (and read from) a plain text files containing the 1-based index of the first data point of each interval.
Once the partition is satisfactory, type
y or click
on the middle button to proceed to the film decay estimation.
at this point cancels the process.
Next, each interval is successively displayed full scale; for each interval, a text message is displayed to tell you what's on:
Film loss estimation on interval 46.80 -> 91.80 ( 2/ 4) (0, mid) : no film loss (k) : enter manually k (l) : linear regression estimation
If you think that film loss is negligible on this interval, type
0 or click the middle button. Else, typing
l lets you
select the X-range that is used for fitting the data to a straight
line (like with the reg command) and measuring the time constant of the decay for the
After all intervals have been examined, SOAS exits the command replacing the buffer with the normalized time-dependent coverage. Use the div command to divide the raw chronoamperometric data by this coverage to correct the signal for film loss.
Christophe Leger 2009-02-24