Description¶
- Class:
- Alias:
ramp_fit
This step determines the mean count rate, in units of counts per second, for each pixel by performing a linear fit to the data in the input file. The fit is done using the “ordinary least squares” method. The fit is performed independently for each pixel. There can be up to three output files created by the step:
The primary output file (“rate”) contains slope and other results at each pixel averaged over all integrations in the exposure.
The secondary product (“rateints”) contains slope and other results for each integration, stored as data cubes.
A third, and optional, output product is also available, containing detailed fit information for each ramp segment for each pixel.
The three types of output products are described in more detail below.
The count rate for each pixel is determined by a linear fit to the cosmic-ray-free and saturation-free ramp intervals for each pixel. Hereafter such intervals will be referred to as a “segment.” The fitting algorithm uses an ‘optimal’ weighting scheme, as described by Fixsen et al, PASP, 112, 1350. Details of the computations are given below.
Segments are determined using the 4-D GROUPDQ array of the input data set, under the assumption that the saturation detection and jump detection steps have already been applied, in order to flag occurrences of both saturation and cosmic-ray (CR) hits. Segments are terminated where saturation flags are found. Pixels are processed simultaneously in blocks using the array-based functionality of numpy. The size of the block depends on the image size and the number of groups per integration.
Upon successful completion of this step, the status keyword S_RAMP will be set to “COMPLETE”.
Note that the core algorithms for this step are called from the external package
stcal, an STScI effort to unify common calibration processing algorithms
for use by multiple observatories.
Multiprocessing¶
This step has the option of running in multiprocessing mode. In that mode it will split the input data cube into a number of row slices based on the number of available cores on the host computer and the value of the max_cores input parameter. By default the step runs on a single processor. At the other extreme if max_cores is set to ‘all’, it will use all available cores (real and virtual). Testing has shown a reduction in the elapsed time for the step proportional to the number of real cores used. Using the virtual cores also reduces the elapsed time but at a slightly lower rate than the real cores.
Detailed Algorithms¶
Special Cases¶
If the input dataset has only a single group in each integration, the count rate for all unsaturated pixels in that integration will be calculated as the value of the science data in that group divided by the group time. If the input dataset has only two groups per integration, the count rate for all unsaturated pixels in each integration will be calculated using the differences between the two valid groups of the science data.
For datasets having more than a single group in each integration, a ramp having a segment with only a single group is processed differently depending on the number and size of the other segments in the ramp. If a ramp has only one segment and that segment contains a single group, the count rate will be calculated to be the value of the science data in that group divided by the group time. If a ramp has a segment having a single group, and at least one other segment having more than one good group, only data from the segment(s) having more than a single good group will be used to calculate the count rate.
The data are checked for ramps in which there is good data in the first group, but all first differences for the ramp are undefined because the remainder of the groups are either saturated or affected by cosmic rays. For such ramps, the first differences will be set to equal the data in the first group. The first difference is used to estimate the slope of the ramp, as explained in the ‘segment-specific computations’ section below.
If any input dataset contains ramps saturated in their second group, the count rates for those pixels in that integration will be calculated as the value of the science data in the first group divided by the group time.
MIRI First and Last Frames¶
The MIRI first frame correction step flags all pixels in the first group of each integration, so that those data do not get used in either the jump detection or ramp fitting steps. Similarly, the MIRI last frame correction step flags all pixels in the last group of each integration. The ramp fitting will only fit data if there are at least 2 good groups of data and will log a warning otherwise.
NIRCam Frame 0¶
If the input data contains a frame zero data cube, those data will be used to estimate a slope for pixels that are saturated in all groups. If all groups in an integration are flagged as SATURATED for a given pixel, the frame zero data array is examined to determine whether or not it is also saturated. Saturated elements of the frame zero array are set to zero by the preceding saturation step in the pipeline. Unsaturated elements will have non-zero values in the frame zero array. If the frame zero is not saturated, then it’s value will be divided by the frame time for the exposure in order to compute a slope for the pixel in that integration. This is analagous to the situation in which only the first group in an integration is unsaturated and used by itself to compute a slope (see above).
Note that the computation of slopes from either a single group or single frame zero
value is disabled when the step parameter suppress_one_group is set to True.
In this case the slope value for a pixel with only one good sample will be set to zero.
All Cases¶
For all input datasets, including the special cases described above, arrays for the primary output (rate) product are computed as follows.
After computing the slopes for all segments for a given pixel, the final slope is determined as a weighted average from all segments in all integrations, and is written as the primary output product. In this output product, the 4-D GROUPDQ from all integrations is collapsed into 2-D, merged (using a bitwise OR) with the input 2-D PIXELDQ, and stored as a 2-D DQ array. The 3-D VAR_POISSON and VAR_RNOISE arrays from all integrations are averaged into corresponding 2-D output arrays. In cases where the median rate for a pixel is negative, the VAR_POISSON is set to zero, in order to avoid the unphysical situation of having a negative variance.
The slope images for each integration are stored as a data cube in a second output data product (rateints). Each plane of the 3-D SCI, ERR, DQ, VAR_POISSON, and VAR_RNOISE arrays in this product corresponds to the result for a given integration. In this output product, the GROUPDQ data for a given integration is collapsed into 2-D, which is then merged with the input 2-D PIXELDQ to create the output DQ array for each integration. The 3-D VAR_POISSON and VAR_RNOISE arrays are calculated by averaging over the fit segments in the corresponding 4-D variance arrays.
A third, optional output product is also available and is produced only when the step parameter ‘save_opt’ is True (the default is False). This optional product contains 4-D arrays called SLOPE, SIGSLOPE, YINT, SIGYINT, WEIGHTS, VAR_POISSON, and VAR_RNOISE that contain the slopes, uncertainties in the slopes, y-intercept, uncertainty in the y-intercept, fitting weights, the variance of the slope due to poisson noise only, and the variance of the slope due to read noise only for each segment of each pixel, respectively. The y-intercept refers to the result of the fit at an effective exposure time of zero. This product also contains a 3-D array called PEDESTAL, which gives the signal at zero exposure time for each pixel, and the 4-D CRMAG array, which contains the magnitude of each group that was flagged as having a CR hit. By default, the name of this output file will have the suffix “_fitopt”. In this optional output product, the pedestal array is calculated for each integration by extrapolating the final slope (the weighted average of the slopes of all ramp segments in the integration) for each pixel from its value at the first group to an exposure time of zero. Any pixel that is saturated on the first group is given a pedestal value of 0. Before compression, the cosmic ray magnitude array is equivalent to the input SCI array but with the only nonzero values being those whose pixel locations are flagged in the input GROUPDQ as cosmic ray hits. The array is compressed, removing all groups in which all the values are 0 for pixels having at least one group with a non-zero magnitude. The order of the cosmic rays within the ramp is preserved.
Slope and Variance Calculations¶
Slopes and their variances are calculated for each segment, for each integration, and for the entire exposure. As defined above, a segment is a set of contiguous groups where none of the groups is saturated or cosmic-ray impacted. The appropriate slopes and variances are output to the primary output product, the integration-specific output product, and the optional output product. The following is a description of these computations. The notation in the equations is the following: the type of noise (when appropriate) will appear as the superscript ‘R’, ‘P’, or ‘C’ for readnoise, Poisson noise, or combined, respectively; and the form of the data will appear as the subscript: ‘s’, ‘i’, ‘o’ for segment, integration, or overall (for the entire dataset), respectively.
Optimal Weighting Algorithm¶
The slope of each segment is calculated using the least-squares method with optimal weighting, as described by Fixsen et al. 2000, PASP, 112, 1350; Regan 2007, JWST-STScI-001212. Optimal weighting determines the relative weighting of each sample when calculating the least-squares fit to the ramp. When the data have low signal-to-noise ratio \(S\), the data are read noise dominated and equal weighting of samples is the best approach. In the high signal-to-noise regime, data are Poisson-noise dominated and the least-squares fit is calculated with the first and last samples. In most practical cases, the data will fall somewhere in between, where the weighting is scaled between the two extremes.
The signal-to-noise ratio \(S\) used for weighting selection is calculated from the last sample as:
The weighting for a sample \(i\) is given as:
where \(i_{midpoint}\) is the the sample number of the midpoint of the sequence, and \(P\) is the exponent applied to weights, determined by the value of \(S\). Fixsen et al. 2000 found that defining a small number of P values to apply to values of S was sufficient; they are given as:
Minimum S |
Maximum S |
P |
|---|---|---|
0 |
5 |
0 |
5 |
10 |
0.4 |
10 |
20 |
1 |
20 |
50 |
3 |
50 |
100 |
6 |
100 |
10 |
Segment-specific Computations¶
The variance of the slope of a segment due to read noise is:
where \(R\) is the noise in the difference between 2 frames, \(ngroups_{s}\) is the number of groups in the segment, and \(tgroup\) is the group time in seconds (from the keyword TGROUP). The derivation of this equation is given in the appendix of this section, at readnoise variance derivation..
The variance of the slope in a segment due to Poisson noise is:
where \(gain\) is the gain for the pixel (from the GAIN reference file), in e/DN. The \(slope_{est}\) is an overall estimated slope of the pixel, calculated by taking the median of the first differences of the groups that are unaffected by saturation and cosmic rays, in all integrations. This is a more robust estimate of the slope than the segment-specific slope, which may be noisy for short segments.
The combined variance of the slope of a segment is the sum of the variances:
Integration-specific Computations¶
The variance of the slope for an integration due to read noise is:
where the sum is over all segments in the integration.
The variance of the slope for an integration due to Poisson noise is:
The combined variance of the slope for an integration due to both Poisson and read noise is:
The slope for an integration depends on the slope and the combined variance of each segment’s slope:
Exposure-level Computations¶
The variance of the slope due to read noise depends on a sum over all integrations:
The variance of the slope due to Poisson noise is:
The combined variance of the slope is the sum of the variances:
The square-root of the combined variance is stored in the ERR array of the primary output.
The overall slope depends on the slope and the combined variance of the slope of each integration’s segments, so is a sum over integrations and segments:
Variances in Output Products¶
If the user requests creation of the optional output product, the variances of segment-specific slopes due to Poisson noise, \(var^P_{s}\), and read noise, \(var^R_{s}\), are stored in the VAR_POISSON and VAR_RNOISE file extensions, respectively.
At the integration-level, the variance of the per-integration slope due to Poisson noise, \(var^P_{i}\), is written to the VAR_POISSON extension of the per-integration (“rateints”) product, and the variance of the per-integration slope due to read noise, \(var^R_{i}\), is written to the VAR_RNOISE extension. The square-root of the combined variance per integration due to both Poisson and read noise, \(var^C_{i}\), is written to the ERR extension.
For the primary exposure-level (“rate”) product, the overall variance in slope due to Poisson noise, \(var^P_{o}\), is stored in the VAR_POISSON extension, the variance due to read noise, \(var^R_{o}\), is stored in the VAR_RNOISE extension, and the square-root of the combined variance, \(var^C_{o}\), is stored in the ERR extension.
Weighted Readnoise Variance¶
If the charge migration
step has been performed prior to ramp fitting, any group whose value exceeds the
signal_threshold parameter will have been flagged with the CHARGELOSS and DO_NOT_USE
data quality flags. Due to the DO_NOT_USE flags, such groups will be excluded
from the slope calculations.
However, it is desired to have a readnoise variance that is similar to pixels unaffected by charge migration, so an additional type of variance will be calculated, in which the excluded groups mentioned above will be included. This additional, ‘weighted’, readnoise variance is used for weighting in the resample step later in the pipeline. The ‘weighted’ readnoise variance is written to the VAR_RNOISE extension of each of the 3 output products.
The original (‘conventional’) type of readnoise variance described earlier is still used internally in other variance calculations but, as mentioned above, is no longer written to the separate variance extension.