Not Processable exposures

What happens if the ST-ECF association pipeline cannot find accurate pointing information for a given exposure ?

Two possibilities:

Or the jitter files are missing, or empty
or the jitter files are available but they don't provide so precise pointing information

In the first case the exposure cannot be part of any association, since the pointing information is missing.

In the second case, while pointing information is available, it is not possible to reliably compute the offsets of that exposure with respect to the others in the association.

The last case is classified as "G" (for Groupable, since it can be part of the association [we know were HST was pointing, roughly at least]), as opposed to the "P" flag issued for exposures with reliable pointing information (Processable, [not only it is groupable, but the shifts are so reliably computed that an automatic pipeline can process it]).

Hence, being the shifts from the other members of the association unknown, "G" exposures won't be used by the WFPC2 reduction pipeline (the one producing the cosmic ray-free images).

In details, an exposure is marked as "G" if at least one of the following conditions holds:

GUIDEACT != "FINELOCK" St. Dev. (Right Ascension) * cos(Declination) > 5 mas St. Dev. (Declination) > 5 mas St. Dev. (Roll angle) > 5 arcsec St. Dev. for RA, DEC or Roll have (for any reason) improper (NULL) values lockloss > 0.

Statistics

Number of "G" exposures / total number of exposures

Exposures Statistics (as of 9 Dec 1997)
Member Flag G	6323
Member Flag P	10308
Member flag unclassified (single exposures)	4873
Total No. of Exposures part of Associations	16631
Total No. of Exposures	21504

Guiding Mode Statistics

Mainly the member flag is set to "G" because of a not accurate guiding mode :

Guiding modes and pointing accuracy
Guiding Mode	Abs. Accuracy	Rel. Accuracy
2 FGS Fine Lock Guiding	0.5 ÷ 2 arcsec	1 ÷ 50 mas
GYRO Control	2 ÷ 100 arcsec	1÷ 5 mas per second of obs. time (drift)
Single FGS Pitch and Yaw + GYRO Roll Control	0.5 ÷ 5 arcsec	1 ÷ 5 mas per second of obs. time (drift)
COARSETRACK	-	-

GUIDEACT Statistics (over 21504 exposures)
NULL 1281

?? 5

COARSETRACK 49

COARSETRACK/G 117

FINELOCK/GYRO 4520

GYRO 1031

FINELOCK 14501

**GUIDEACT Statistics** (over 21504 exposures)
NULL	1281
??	5
COARSETRACK	49
COARSETRACK/G	117
FINELOCK/GYRO	4520
GYRO	1031
FINELOCK	14501

GUIDEACT Members Statistics
(over the 16631 exposures part of an association)
NULL 949

?? 2

COARSETRACK 36

COARSETRACK/G 73

FINELOCK/GYRO 3339

GYRO 724

FINELOCK 11508

**GUIDEACT Members Statistics**
(over the 16631 exposures part of an association)
NULL	949
??	2
COARSETRACK	36
COARSETRACK/G	73
FINELOCK/GYRO	3339
GYRO	724
FINELOCK	11508

Summarizing, there are a total of 7004 (as of 9 Dec 1997) exposures with a guiding mode different from "FINELOCK". In the following table I have distinguished between Single (for single exposures, not part of any association) and Member for those exposures which were found part of an association.

**GUIDEACT Statistics Summary**
Single	NOT FINELOCK	1881
Member	NOT FINELOCK	5123

Single	FINELOCK	2992
Member	FINELOCK	11508

Total	NOT FINELOCK	7003
Total	FINELOCK	14501

Other reasons for "NOT PROCESSABLE" flag

It could be that even when in 2 FGS FINELOCK guiding mode the observation has been flagged as "G". There are a number of possible reasons for this to happen:

FINELOCK but "G" members

No of "G" members having FINELOCK	1200
lockloss > 0	13
ra_sd > 5	870 (-676?)
dec_sd > 5	143
ra_sd && dec_sd > 5	170
roll_sd > 5000	4