pandas - alternative to recursive sql / dedupe columns -

-

hi have data set 1 many relations. want reduce 1-1 concatenating distinct values each column

here example:

i have data set called customer , product affiliation (pa). 1 customer can have multiple pa different times.

here input:

╔════════╦═══════╦══════╦══════╗ ║ cust   ║pa1    ║ pa2  ║ pa3  ║ ╠════════╬═══════╬══════╬══════╣ ║      ║ h     ║ m    ║ l    ║ ║      ║ h     ║ l    ║ m    ║ ║      ║ h     ║ m    ║ h    ║ ╚════════╩═══════╩══════╩══════╝

desired output:

╔════════╦══════╦══════════╦═══════╗ ║  cust  ║  pa1 ║   pa2    ║  pa3  ║ ╠════════╬══════╬══════════╬═══════╣ ║      ║  h   ║   m&l    ║ l&m&h ║ ╚════════╩══════╩══════════╩═══════╝

i have multiple pa columns. want know if there generic python panda code can apply thank minc

here's expanded version of sample data make more clear how code operates on groups rather whole dataframe:

df = pd.dataframe({ 'cust':list('aaabbb'),'pa1':list('hhhmmh'),                     'pa2': list('mlmmmm'),'pa3':[1,2,3,3,3,1] })    cust pa1 pa2 pa3 0      h   m   1 1      h   l   2 2      h   m   3 3    b   m   m   3 4    b   m   m   3 5    b   h   m   1

the first step create new dataframe 1 row per customer. quick , dirty way fine example @ hand might want differently depending on want other columns.

df2 = df.groupby('cust').sum()        pa1  pa2  pa3 cust                    hhh  mlm  123 b     mmh  mmm  331

with template desired output, can loop on relevant columns of original dataframe , use groupby/apply join/unique replace values in existing columns:

for col in ['pa1','pa2','pa3']:     df2[col] = df.groupby('cust')[col].apply( lambda x:                                                '&'.join( x.astype(str).unique() ))        pa1  pa2    pa3 cust                        h  m&l  1&2&3 b     m&h    m    3&1

Comments

Post a Comment

Popular posts from this blog

python - jinja2: TemplateSyntaxError: expected token ',', got 'string' -

-
i new jinja2 , having issue using python regular expression (re). in following code bold lines have error string in them. {% block content %} <div class="container"> {% l in lines %} {% if re.search(r"error", l) %} {# <<< throws error #} <b> {{ l }} </b> {% else %} {{ l }} <hr> {% endif %} {% endfor %} </div> {% endblock %} the re.search above throws following error: jinja2.exceptions.templatesyntaxerror templatesyntaxerror: expected token ',', got 'string' raw python code not supported in jinja2 template syntax. {% if re.search(r"error", l) %} replace line with {% if "error" in l %} can fix problem. if logical condition more complicated, should consider defining own custom filters (which can call python code) or complicated things in view layer. go check global namespace .
Read more

node.js - NodeJS remote terminal to Dropbear OpenWRT-Server -

-
i wandering whether there valid solution have remote terminal openwrt-box out of nodejs-app? connecting terminal works: ssh -i ~/.mykeys/id_rsa root@192.168.178.39 busybox v1.23.2 (2015-04-22 23:25:48 utc) built-in shell (ash) root@openwrt:~# the only interactive ssh solution nodejs doesn't interactive part described in readme.md following: var client = require('ssh2').client; var conn = new client(); conn.on('ready', function() { console.log('client :: ready'); conn.shell(function(err, stream) { if (err) throw err; stream.on('close', function() { console.log('stream :: close'); conn.end(); }).on('data', function(data) { console.log('stdout: ' + data); }).stderr.on('data', function(data) { console.log('stderr: ' + data); }); stream.end('ls -l\nexit\n'); }); }).connect({ host: '192.168.100.100', port: 22, username...
Read more

Qt4: how to send QString inside a struct via QSharedMemory -

-
i have struct struct control_data{ int column_number; qstring cell; }; i need send thread of qsharememory. read can't because qstring contains pointers inside. other ways? you have serialize struct byte array. can convert qstring const char* this: mystring.tostdstring().c_str(); but serializing qstring should work. the first step serialize struct qdatastream using qt, example here . then once struct can read , written can pass shared memory. a complete example of using qsharedmemory can found here . here relevant code: // first, test whether shared memory segment attached process. // if so, detach if (sharedmem.isattached()) { sharedmem.detach(); } ... qbuffer buffer; buffer.open( qbuffer::readwrite ); qdatastream out( &buffer ); out << youstruct; int size = buffer.size(); // size of int + size of qstring in bytes if ( !sharedmem.create( size ) ) { return; } // write shared memory sharedmem.lock(); char *to = (char*...
Read more