Opening a .DAT file

I need to open a .DAT file that was created in 1988. This .DAT file was created by Fortran, which was programmed to collect streaming data.

The first ten rows are a header and I believe that each and every row is 140 bytes.

I have tried several programs and I get garbage. When I tried Matlab, I get actual numbers, but no header (which probably means it is all garbage).

I am using the "fopen" and "fread" commands in matlab.

Can anyone give me some insight on how to open this file correctly?

I can send the file upon request.

PS - I am a fairly new user, so cut me some slack.

This is what I get when I open it with Notepad:

14ma30 pinsfx s20e25 w2c p vel v D € 14MA30PINSFX314:13 VAG” F²FRp РÐB Ø@ÃõÀ¶ó]AòÒ æ=ܟýÌgš;€P3>sªÁ;@Ú‘<àuk;P¦> v< ½%¾s‹ =¨Ð†¼`¾’¶½ÌO­¼àTŽ=mñ¾ˆèZ<Àb½í~¾ÐÝL½Àž85†<€({½P:“½ƒE">8¼ •=î‡>ˆ§s=P8>[²¼Àßf=5 €@ €@ €@ €@ €@ A A A A A A E € B ÈB E @ÞB  B B ðB 0=#lî>†p¿Ët±»J4s;Æa´;t@ÚÿͼԨ¬?ƪ1¿÷~9»mù°»´jO<hœ@˜ÿ®¼bL@/I¿ u¼l¯¼u«¨<Û@ºý>Aƒ@µKG¿õª;N€/¼°/È<mZ@|þ`=à­”@wî,¿÷ɼžÌ)¼Þ×-<®¨@Nþ0=‡pî>ip¿ t²»ù¬t;Bk³;“\@Øÿͼš‡¬?«1¿ýz'»ÎÀ®»c²Q<åÄ@˜ÿ­¼ô¸@~I¿tq¼gô°¼Ê*©< J@ºý>C¥ƒ@ýKG¿ñ£;½ê-¼ïvÆ<E@„þa=u”@™î,¿ññƼ ò(¼N.<ì{@Xþ0=étî>Lp¿ut´»x)u;*r²;Ô«@Øÿͼff¬?F«1¿w»Œ{¬»øS<í@–ÿ­¼ˆ%@ÍI¿Þn¼Lƒ°¼š®©<¸@¼ý>Cºƒ@ELG¿ í;ýO+¼§»Ã<Ûä@Žþa= ”@»î,¿Øìļ©&¼ÁQ/<ºP@`þ0=Myî>/p¿Jtµ»ý£v;·~±;¯ú@Øÿͼ=E¬?†«1¿s»L1«»{9V<¯@–ÿ¬¼4’@I¿Dk¼Q±¼û7ª<M @¼ý>?σ@ŒLG¿ é–;P¨)¼ øÁ<€ª@–þa=˜ÿ”@Ýî,¿ÃçÁ¼[?$¼ÏŽ0<Ð&@jþ0=°}î>p¿t·»òw;Jˆ±;™I@Øÿͼ'$¬?Æ«1¿ oẉ¦©»‡hX<NB@”ÿ«¼ÿ@gI¿ª g¼´Ä±¼ÚÁª<p~@¼ý>7äƒ@ÓLG¿å;Pú(¼J*¿<Ýo@ þb=%p”@ÿî,¿­â¾¼ge"¼•Ë0<Qú@tþ0=‚î>öo¿õs¸»‡–x;•°;Q˜@Øÿͼ,¬?¬1¿ k¼ºŸÂ§»š‰Z<p@”ÿ«¼Jl@±I¿ d¼Tt²¼çKª<vÌ@¾ý>$ùƒ@MG¿ûà‰;;&¼ZP½<5@¨þb=¬à”@#ï,¿•Ý»¼¼!¼ð1<DÍ@|þ0=w†î>Ùo¿Ésº»*y;嚯;jç@Øÿ̼Xâ¬?F¬1¿ g—º¡¥»?—\</¡@’ÿª¼ÌÙ@õI¿s a¼;.²¼mÒ«<e@¾ý>ƒ@bMG¿íÜ‚;r$¼Älº<m÷@²þc=/Q”@Dï,¿Ø¸¼ µ¼¤E2<8¢@†þ0=ÚŠî>»o¿žs»»‹z;(¥¯;6@Øÿ̼¸Á¬?‡¬1¿cdºñ~£»¾Ž^<UÖ@’ÿª¼·G@6I¿Ú]¼@î³¼«Q«<Ç"@Àý>Ü"ƒ@ªMG¿ÙØw;«/"¼<~¸<·@ºþc=ªÁ”@fï,¿gÓµ¼‰Ý¼‹‚3<¦v@Žþ0=?î>žo¿ss½»6{;l®®;(…@Øÿ̼J¡¬?ɬ1¿ý^ºò'¡»¡|a<R@ÿ©¼2¶@qI¿CZ¼ò²³¼ Ç«<Â%@Àý>£7ƒ@òMG¿½Ôi;Gd ¼ª„¶<¾v@Âþd=2”@‰ï,¿Qγ¼V¼F¾4< G@–þ0=£“î>o¿Gs¾»µ~|;·­;Ô@Øÿ̼ ¬? ­1¿òZš¹ñDŸ»dc<¢M@Žÿ©¼¸%@§I¿°W¼Gp´¼Ñ(«< @Âý>ZLƒ@8NG¿—Ð[;ú–¼Í‚´<2@Êþd=‡¢”@«ï,¿8É°¼¾-¼›ù5<@žþ0=˜î>co¿s¿»,û};æÀ­;Â"@Øÿ̼@a¬?P­1¿ÞV¶B»žJe<ÌŽ@Œÿ¨¼À–@ÛI¿$T¼!´¼…qª<Ã×@Æý>ÿ`ƒ@€NG¿jÌM;(½¼ÿz±<úæ@Ôþe=ë”@Íï,¿ Ä­¼±U¼Ý35<óë@¦þ0=kœî>Fo¿òrÁ»q~;Yͬ;¹r@Öÿ̼®A¬?–­1¿ÈR™9þ^›»Cg<M×@Šÿ¨¼˜ @ I¿žP¼¿´¼" ª<*‡@Èý>’uƒ@ÇNG¿3È?; â¼’o¯<Eš@Üþe=Bƒ”@ïï,¿¿ª¼K~¼ôm6<¾@®þ0=Рî>)o¿Ær»rï;Þ«;}Á@Öÿ̼c"¬?Þ­1¿­N:¹t™»wOj<¶#@ˆÿ§¼~@8I¿ M¼OA´¼û«ª<u@Ìý>Šƒ@OG¿óÃ2;ð¼ƒ_­<ÈI@äþe=ó”@ð,¿í¹§¼n§¼¡§7<AŽ@¶þ0=7¥î> o¿›rÄ»†i€;Fs«;@Öÿ̼\¬?'®1¿ŽJh:¢©—»Apl<¤x@†ÿ¦¼¢õ@dH¿®ÿI¼š´¼EŒ©<nš@Ðý>ˆžƒ@UOG¿«¿$;S¼ÄKª<›ó@ìþf=×c”@4ð,¿Ó´¤¼mϼfà8<Õ_@¾þ0=›©î>în¿qrÅ»õá€;ï÷ª;ä`@Öÿ˼‘ä¬?s®1¿kF›:| •»êŸn<ÛÔ@„ÿ¦¼5o@H¿GþE¼—´¼g<©<¸@Öý>벃@OG¿[»;­Ÿ¼¿1¨<{—@ôþf=Ô”@Vð,¿¸¯¡¼1ù¼{9<Ä-@Æþ0= ®î>Ñn¿FrÇ»U`;c~©;2¯@Öÿ˼óŬ?î1¿FBÃ:)“»tÐq<);@‚ÿ¥¼wë@ºH¿ïüA¼"¸³¼'¼¨<µW@Üý>?ǃ@äOG¿·;ƒý¼Ä¦<{7@üþg=DD”@yð,¿œªŸ¼d$¼ÙR9<û@Îþ0=g²î>´n¿rÈ»eØ‚;"¨;mÿ@Öÿ˼y§¬?¯1¿>ê:¢‘»·÷s<ϯ@€ÿ¥¼¦j@æH¿Ÿû=¼i~³¼§<»›@äý>…Ûƒ@*PG¿š²ö:>ʼ…ã£<ÜÑ@ÿg=n´”@›ð,¿€¥œ¼áO ¼ÂŒ:<„Ç@Öþ0=жî>—n¿ïqÊ»W‚;|Š¨;žN@Öÿ˼ÿˆ¬?i¯1¿ô9 ;&»xv<j/@|ÿ¤¼Ûì@H¿Wú9¼²²¼ç9¦<Ð@êý>¿ïƒ@rPG¿+®Û:Ç  ¼©¡<Ég@ ÿh=$”@½ð,¿d ™¼Ó| ¼ Ç;<Ê”@Üþ0=6»î>{n¿ÄqË» σ;†§;Ÿ@Öÿ˼†j¬?¯1¿Ê5;üŽ»´x<]¼@zÿ¤¼Cr@CH¿ù4¼2¡±¼ÚF¥<.ö@ôý >îƒ@¹PG¿²©À:ˆi ¼³^ž<²ü@ÿh=¥””@ßð,¿G›–¼ƒ« ¼&<<ï_@äþ0= ¿î>\n¿™qÍ»{Iƒ;R”¦;Ží@Ôÿ˼çK¬?°1¿¢10;²áŒ»({<^R@vÿ£¼û@uH¿Ë÷0¼7 °¼Æ:¥<Z@üý >ƒ@ QG¿.¥¥:Gr ¼`œ<Ú‹@ÿi=±”@ñ,¿,–“¼›Û¼_>=<w(@êþ0= Äî>?n¿nqλUÇ„;Ó¦;À=@Ôÿ˼-¬?}°1¿z-D;ï­Š».

[Ken Atwell]

Posting the first 20 lines (or so) of the data file would help a lot. .DAT is a generic file extension and it can mean just about anything.

[Jake]

This is what I get when I open it with Notepad.

[Walter Roberson]

http://www.mathworks.com/matlabcentral/answers/7924-where-can-i-upload-images-and-files-for-use-on-matlab-answers

Answer 1

[Ken Atwell]

Okay, binary data. :( Do you know what the file is supposed to contain? What is created on a Intel x86 machine (that is, is the floating point format likely to be compatible with your current machine)? Unless you know something about the data you are reading in, you're flying blind.

I might be tempted to try:

 f = fopen('foo.dat, 'rb');
 D = fread(f, inf, '*float');  % Try '*double' too
 fclose(f);
 plot(D);

Does a signal eventually emerge in the plot? If there is a header following by a stream of vector data, you might see a burst of noise (the header) followed by the data.

Beyond this, I don't have a lot of ideas, sorry.

[Walter Roberson]

Definitely a header on there!

[Jake]

This is what I know: For each segment (5 segments), seven pieces of data were stored (x, y, z, a, b, c, d), so records should be nseg*7*4 (for R*4 data) bytes long, which should be 140 bytes by my calculation. The files were opened with fixed length records, so even the header lines are 140 bytes. The header should be 10 lines long.

Does this information help?

With your suggested code, I end up with a 22400 x 1 matrix and it plots fine.

I think I should end with either 1) a string of data the self contains 35 data points before starting a new row, or 2) 35 separate data points before starting a new row.

I appreciate your help. I finally feel like I am getting somewhere.

[Walter Roberson]

Once you have fopen() the file as binary, you can use fseek() to move forward a particular number of bytes.

Question: was the data possibly written in VAX-G floating point format? Support for reading that was dropped a few years ago.

[Jake]

The Data was in written in VAX-D floating point format. I found this forum page that includes a function that will convert VAX-G and VAX-D into 'ieee'. Thanks for the idea and the help. -Jake

http://www.mathworks.com/matlabcentral/fileexchange/22675