Notice that the input data smv.bin was provided twice, but one image was given a scale of zero.ĬCP4-formatted electron density maps. Writing photons.bin as a 0-byte header with 9437184 4-byte floats Header = 0 bytes pixel zero offset = 40 (0 in output) float_add smv.bin smv.bin -scale1 0.625 -scale2 0 -ignore 0 -ignore 65535 -offset 40 -outoffset 0 -outfile photons.bin Mean = 43.2971 rms = 55.748 rmsd = 35.1169 skewness = 376.905 kurtosis = 242337Īlso, the valid pixels in this image have an offset of 40 pixel units, and one photon hitĬreates an average pixel increment of 1.6 levels (ADU). Writing output.bin as a 0-byte header with 9437184 4-byte floatsīut these statistics include the "invalid" pixels (zero and overflows). Header = 0 bytes pixel zero offset = 0 (0 in output) Now we can use float_add to get some statistics : To convert the X-ray image into float values This is why I implemented automatic outlier rejection The yx lsq scale and offset is the scale and offsetįor more complicated cases, outliers can throw off the lsq fit, as well as the You can see that the xy lsq scale reflects the scale and offset we justĪpplied to the map. float_add -header 1104 3Dball.map rescaled.map float_add -header 1104 3Dball.map -scale1 1.234 -outoffset 0.1234 -outfile rescaled.map For example, let's scale and offset this map: It will also give you the correlation coefficient (CC),Īnd the linear least-squares coefficents that will optimally scale one map This gives you all of the stats that the CCP4 mapdump program might give you, Writing output.bin as a 1104-byte header with 32768 4-byte floats Header = 1104 bytes pixel zero offset = 0 (0 in output) This file is float data with a header of 1104 bytes. Now say you want the skewness of an electron density map, such as the one we created in the example forįloatgen, called 3Dball.map. There are no dependencies beyond the standard C It can also generate data to directly replace the voxel values in an electron density map. These flat floating-point files can then be input into image generation programs like Or simply re-scaling a set of floating-point data. The correlation coefficient and least-squares linear fit, before and after applyingįloat_add is useful for direct side-by-side comparision of real and simulated data,įor combining different simulations together (such as spots from This simple program reads two flat files of 4-byte single precision "float" numbersĪnd adds them together with different scale factors to create a new flat file ofĤ-byte single-precision "float" numbers. Float_add: combine/compare files of 4-byte floating-point numbers float_add: combine/compare files of 4-byte floating-point numbers
0 kommentar(er)
