robem make zero pif when normalizing, and PPFT Segmentation fault

39 views

Skip to first unread message

Xiao-Wei Song

unread,

May 30, 2012, 7:25:31 AM5/30/12

to auto...@googlegroups.com

Hi, dear auto3dem experts,

Currently, I think I still could not run through auto3dem for our group's data.

For excluding MPI errors, I compile auto3dem with serial in make_all. But still, PPFT could not work with Segmentation fault.

After careful checking, I found about 1/10 images normalized by auto3dem is fully zero, which is verified by imgstats and visually displayed with robem. After I delete these images and corresponding dat files, PPFT still Segmentation fault.

After I change the verbose from -1 to 9 in ppft_input_iter_1, PPFT does not give any more error hint information.

The command line and output is as follows, any idea to help me, thanks!

$ PPFT < /dev/null ppft_input_iter_1

PFTsearch: Model-based PFT analysis

Parallel version run on 1 CPUs

For doc see: http://cryoem.ucsd.edu/programs.shtm

*** input file passed as cmd line argument

Keyword input found:

threedr_mode = 0

quick_omega = 1

input_mode = 2

bin_factor = 2

symmetry = 532

delta_theta = 1.000000

image_projection_compare_method = 3

ctf_mode = 3

ctf_filter_mode = 0

verbose = 9

filter_factor_1 = 0.1000000

filter_factor_2 = 0.000000

pftrad_lo = 1.000000

pftrad_hi = 360.0000

pftrad_step = 1.000000

annulus_low = 206.0000

annulus_high = 263.5500

radius = 0.000000

temperature_factor = 0.000000

resolution_out = 50.00000

resolution_low = 172.2240

resolution_high = 30.00000

units_out = 1

jcut = 1

sigcut = 0.000000

delta_xy = 0.5000000

num_delta_xy = 7.000000

delta_omega = 0.5000000

num_delta_omega = 7.000000

map_pixsiz = 0.000000

mag_cen = 1.000000

mag_step = 0.000000

mag_num = 1

mag_norm = 1

map_dim = 0

max_origin_shift = 0.3000000

ico_spread = 1.000000

minimum = 0.000000

bg_out = 0

lg_out = 0

mp_out = 0

input_mode_cc = 1

phase_weight = 0.000000

model_filename = hollow_map

prj_filename = pft.prjs

pft_filename = pft.pfts

pftrads_filename = ppft_iter_1.rads

pftres1_filename = ppft_iter_1.res1

pftres2_filename = ppft_iter_1.res2

bg_filename = bg.dat

title = pftsearch_title

filt_fac factor set at - 0.1000000

img file OK: ../../pif/1250_fix_norm_apod.pif

img file OK: ../../pif/1698_fix_norm_apod.pif

img file OK: ../../pif/1315_fix_norm_apod.pif

img file OK: ../../pif/1296_fix_norm_apod.pif

img file OK: ../../pif/1191_fix_norm_apod.pif

img file OK: ../../pif/1823_fix_norm_apod.pif

img file OK: ../../pif/1417_fix_norm_apod.pif

img file OK: ../../pif/1552_fix_norm_apod.pif

img file OK: ../../pif/1562_fix_norm_apod.pif

img file OK: ../../pif/1771_fix_norm_apod.pif

img file OK: ../../pif/1490_fix_norm_apod.pif

img file OK: ../../pif/1316_fix_norm_apod.pif

img file OK: ../../pif/1833_fix_norm_apod.pif

img file OK: ../../pif/1088_fix_norm_apod.pif

img file OK: ../../pif/1714_fix_norm_apod.pif

PFT parameters

---------------------------------------------

bin factor 2

symmetry 532

# model views/asu 370

fft_dim1,fft_dim2 1024 1024

res_lo,res_hi 172.224 30.000

res_min,res_max 14.222 81.647

pixres_min/max 144.000 25.084

pixel size 1.196

nannuli 180

pftrad_lo,pftrad_hi 1.000 180.000

ann_lo,ann_hi 103 131

nrot 2048

delta omega 0.176

model xcen,ycen 181.000 181.000

mag(_cen,_step,_num) 1.000 0.000 1

mag range(lo,hi) 1.000 1.000

----------------------------------------------

Number of parameter files = 15

Number of image files = 15

Total number of images: 150

param # 1 orient file: 1250.dat_000

image # 1 image file: ../../pif/1250_fix_norm_apod.pif

views = 8

numcols = 721 numrows = 721

Ptle records = 8

param # 2 orient file: 1698.dat_000

image # 2 image file: ../../pif/1698_fix_norm_apod.pif

views = 11

numcols = 721 numrows = 721

Ptle records = 11

param # 3 orient file: 1315.dat_000

image # 3 image file: ../../pif/1315_fix_norm_apod.pif

views = 8

numcols = 721 numrows = 721

Ptle records = 8

param # 4 orient file: 1296.dat_000

image # 4 image file: ../../pif/1296_fix_norm_apod.pif

views = 9

numcols = 721 numrows = 721

Ptle records = 9

param # 5 orient file: 1191.dat_000

image # 5 image file: ../../pif/1191_fix_norm_apod.pif

views = 6

numcols = 721 numrows = 721

Ptle records = 6

param # 6 orient file: 1823.dat_000

image # 6 image file: ../../pif/1823_fix_norm_apod.pif

views = 8

numcols = 721 numrows = 721

Ptle records = 8

param # 7 orient file: 1417.dat_000

image # 7 image file: ../../pif/1417_fix_norm_apod.pif

views = 12

numcols = 721 numrows = 721

Ptle records = 12

param # 8 orient file: 1552.dat_000

image # 8 image file: ../../pif/1552_fix_norm_apod.pif

views = 11

numcols = 721 numrows = 721

Ptle records = 11

param # 9 orient file: 1562.dat_000

image # 9 image file: ../../pif/1562_fix_norm_apod.pif

views = 17

numcols = 721 numrows = 721

Ptle records = 17

param # 10 orient file: 1771.dat_000

image # 10 image file: ../../pif/1771_fix_norm_apod.pif

views = 10

numcols = 721 numrows = 721

Ptle records = 10

param # 11 orient file: 1490.dat_000

image # 11 image file: ../../pif/1490_fix_norm_apod.pif

views = 10

numcols = 721 numrows = 721

Ptle records = 10

param # 12 orient file: 1316.dat_000

image # 12 image file: ../../pif/1316_fix_norm_apod.pif

views = 8

numcols = 721 numrows = 721

Ptle records = 8

param # 13 orient file: 1833.dat_000

image # 13 image file: ../../pif/1833_fix_norm_apod.pif

views = 13

numcols = 721 numrows = 721

Ptle records = 13

param # 14 orient file: 1088.dat_000

image # 14 image file: ../../pif/1088_fix_norm_apod.pif

views = 16

numcols = 721 numrows = 721

Ptle records = 16

param # 15 orient file: 1714.dat_000

image # 15 image file: ../../pif/1714_fix_norm_apod.pif

views = 10

numcols = 721 numrows = 721

Ptle records = 3

Segmentation fault

Xiaowei

Xiao-Wei Song

unread,

May 30, 2012, 7:34:11 AM5/30/12

to auto...@googlegroups.com

I forget to tell you guys, auto3dem works for P22demo under both serial and parallal version.

But it failed for my images. My images is transformed as the way you helped me:

em2em -t <<eof

IMAGIC

MRC

MULTIPLE_LOCATION_FILE

1000.img

1000.mrc,0,3

YES

eof

bconvert 1000.mrc -datatype int -output mrc

echo '2D' | mrc2pif 1000.mrc 1000.pif

The above programs version are:

1. em2em, EM2EM (vs. 28-Mar-2011)

2. bconvert, 1.8.1-20120516 (64 bit)

3. mrc2pif, v 4.01.12. December 6, 2011

在 2012年5月30日星期三UTC+8下午7时25分31秒，Xiao-Wei Song写道：

在 2012年5月30日星期三UTC+8下午7时25分31秒，Xiao-Wei Song写道：

Giovanni Cardone

unread,

May 30, 2012, 1:30:03 PM5/30/12

to auto...@googlegroups.com

Hi Xiaowei,

usually this kind segmentation fault comes from some inconsistency in the data.
You can check two things:
1. verify that the reference map has the same linear size as the particles. They need to be the same odd size.
2. verify the integrity of the stack specified in the parameter file 1714.dat_000. The program gets two different values (10 and 3) for the number of particles in it.

Regarding the stack of images that appear as zeroed, did you try to open them with other visualization programs? Anyway, I would need a test case, or a procedure that reproduces the problem, in order to see what is going on.

Giovanni

-- 
Giovanni Cardone
Baker's Laboratory
University of California, San Diego
Department of Chemistry and Biochemistry
9500 Gilman Drive, NSB 4107, MC-0378
La Jolla, CA, 92093-0378
(858)534-5852

Reply all

Reply to author

Forward

0 new messages